Thermodynamics Basic Principles

Mechanical Engineering Series
Alan M. Whitman
Thermodynamics:
Basic Principles
and Engineering
Applications
Series Editor
Francis A. Kulacki, Department of Mechanical Engineering, University of
Minnesota, Minneapolis, MN, USA
The Mechanical Engineering Series presents advanced level treatment of topics on
the cutting edge of mechanical engineering. Designed for use by students,
researchers and practicing engineers, the series presents modern developments in
mechanical engineering and its innovative applications in applied mechanics,
bioengineering, dynamic systems and control, energy, energy conversion and
energy systems, fluid mechanics and fluid machinery, heat and mass transfer,
manufacturing science and technology, mechanical design, mechanics of materials,
micro- and nano-science technology, thermal physics, tribology, and vibration and
acoustics. The series features graduate-level texts, professional books, and research
monographs in key engineering science concentrations.
More information about this series at http://www.springer.com/series/1161

Alan M. Whitman
Thermodynamics: Basic
Principles and Engineering
Applications
123
Alan M. Whitman
Department of Mechanical Engineering
Villanova University
Villanova, PA, USA
Additional material to this book can be downloaded from http://extras.springer.com.
ISSN 0941-5122 ISSN 2192-063X (electronic)

ISBN 978-3-030-25220-5 ISBN 978-3-030-25221-2 (eBook)
https://doi.org/10.1007/978-3-030-25221-2
© Springer Nature Switzerland AG 2020
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my wife Rella and to
Karen, Phyllis, Roy, and Noam
Preface
In this text, I have presented thermodynamics as mechanics, fluid mechanics, and

heat transfer are presented, by casting the discussion in terms of familiar mathe-
matical concepts that students have learned previously in three semesters of cal-
culus. I have adopted an axiomatic presentation and have introduced the two laws
of thermodynamics as differential equations for each newly defined variable,
internal energy, and entropy (in the case of entropy two axioms are used to simplify
the presentation). Although this approach differs from the inductive manner in
which the subject developed historically and which is the presentation used in most
existing textbooks, I believe it has many pedagogical advantages; for example, it is
simpler, discussion of the Maxwell relations and other important thermodynamic
results occur naturally in the development of the entropic equation of state, heating,
and power bounds for special systems including the impossibility of perpetual
motion of the second kind and Carnot's result for heat engines, follow deductively
from the two laws, and, most importantly, it makes thermodynamics more palatable
for students, by making learning it like learning all their other engineering science
courses.
Although the treatment is deductive and tends to be mathematical, I have tried to
make the text student friendly by minimizing the level of abstraction, and by
motivating new concepts by relating them to older, more familiar, ones. In dis-
cussing the properties pressure and temperature, I have included material on their
kinetic theory expressions in order to enhance physical intuition. Further on when
introducing new properties of internal energy and entropy, I have also included
kinetic theory material; in these cases, there is even more benefit since these
properties lack the familiarity of the previous ones. I have used a format for
organizing problem statement data that makes clear whether a problem is properly
stated, and how to proceed toward a solution. This format is applicable to all change
of state problems encountered in thermodynamics including those involving both
open and closed systems. This problem-solving methodology is described initially
in Chap. 2 and is continued throughout the text. Every example in the book is
solved using the same format, the idea being that the repetition will make students
comfortable with the process, and better able to use it on their own. Repetition is
vii
viii Preface
also used in discussing equations of state. In Chap. 2, the mechanical equation of

state is written as a differential equation by differentiating vðtÞ ¼ v½TðtÞ; pðtÞ and
noting that the partial derivatives of specific volume with respect to temperature and
pressure are va and vb respectively; the quantity a is the coefficient of thermal
expansion and b is the isothermal compressibility, each of these being an experi-
mentally determinable quantity defined and discussed in Chap. 1. The equation of
state for specific volume is then found by integrating, which results in integrals
whose integrands are known quantities. In Chap. 4, the same approach is applied to
the energetic equation of state uðtÞ ¼ u½TðtÞ; vðtÞ, in this case, the partial deriva-
tives are cv and ðcp cv Þ=ðvaÞ respectively; the quantities cv and cp are the specific
heats at constant volume and at constant pressure, each of these being an experi-
mentally determinable quantity defined and discussed in Chap. 3. The equation of
state for specific internal energy is then obtained in terms of integrals whose
integrands are likewise known quantities. The enthalpic equation of state is also
treated here in the same way. Finally, in Chap. 5, the same approach is again
applied, this time to the entropic equation of state. In all these instances, the
integrals are evaluated in the special cases of liquids and solids, and ideal and
perfect gases, so that all the simple equations of state are derived from first prin-
ciples using relatively simple mathematics that has been learned in previous courses
and reviewed in Chap. 2. This tactic provides a common thread that ties together the
various new concepts that are introduced, as well as a familiar framework into
which they are incorporated. I have also used a matrix format for the property tables
of steam and R-12 that helps students visualize this data. This is, especially, true for
compressed liquid states. In addition, it makes interpolation near the saturation
curve easier to comprehend and to do.
It is necessary to mention a few things about notation. In the attempt to provide a
good notation, one is always faced with the dilemma of having it convey more
information, and risking it being confusing because it conveys too much, or having
it convey less information, and risking it being confusing because it conveys too
little. With this in mind, I have endeavored to make symbols that represent different
things different. The most important case of this is in distinguishing gauge quan-
tities from absolute ones. Thus, gauge pressure is denoted by pg while absolute
pressure is denoted by p. Although this is unremarkable, I have also distinguished
the gauge temperatures Fahrenheit, Tf , and Celsius, Tc , from each other because
they have different zero points, and distinguished them both from absolute tem-
perature, T, as well. Although most texts do not make these distinctions in tem-
perature measures, I believe they are useful for several reasons; first, when using
them, the same conversion equations apply among these quantities as apply to all
other relative quantities such as location and time, (e.g., Tf ¼ Tc þ 32 F) and
second, using them minimizes the common student error of using a Tf or Tc value in
an equation written for T, such as the ideal gas law. I use the same philosophy at
other places in the text, for example, QA=B denotes the heat transfer to body A from
body B; however, in problems involving only one body, where there is little chance
for confusion, I simply use Q to denote the heat transfer to it.
Preface ix
The basic structure of the text has been heavily influenced by reading works on
thermodynamics by C. Truesdale, J. Serrin, C. L. Ericksen, and their colleagues.
Although this book is by no means a text on Rational Thermodynamics, I found that
many of my pedagogical concerns had been already addressed by those authors,
and I believe that I was able to cast much of their thought into a form which is
suitable for a one-semester undergraduate course. Naturally, the responsibility for
using this particular form of presentation is mine alone.
Villanova, PA, USA Alan M. Whitman

April 2019
Contents
1 Measurement and Properties of Matter . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Dimensions and Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Fundamental and Derived Dimensions . . . . . . . . . . . . . . . 6
1.2.2 Absolute and Relative Quantities . . . . . . . . . . . . . . . . . . 9
1.3 Properties of Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.1 Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.2 Weight and Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.3.3 Density and Specific Volume . . . . . . . . . . . . . . . . . . . . . 14
1.3.4 Velocity and Acceleration . . . . . . . . . . . . . . . . . . . . . . . . 15
1.3.5 Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.3.6 Impulse and Momentum . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.3.7 Work and Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.3.8 Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.3.9 Heating, Hotness, and Temperature . . . . . . . . . . . . . . . . . 32
1.3.10 Coefficient of Thermal Expansion . . . . . . . . . . . . . . . . . . 41
1.3.11 Compressibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.2 Thermostatics of Pure Fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.3 The State Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.3.1 The Geometry of Curves . . . . . . . . . . . . . . . . . . . . . . . . 52
2.3.2 The Geometry of Surfaces . . . . . . . . . . . . . . . . . . . . . . . 55
2.3.3 Thermostatic and Thermodynamic Problems . . . . . . . . . . 60
xi
xii Contents
2.4 The Mechanical Equation of State . . . . . . . . . . . . . . . . . . . . . . . . 64

2.4.1 Liquids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.4.2 Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.4.3 Multicomponent Systems . . . . . . . . . . . . . . . . . . . . . . . . 78
2.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3 Work and Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.2 Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.2.1 Conservative Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.2.2 Reversible and Irreversible Work . . . . . . . . . . . . . . . . . . 91
3.2.3 Continuous Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.2.4 External Determination of Work Done . . . . . . . . . . . . . . 94
3.2.5 Internal Determination of Work Done . . . . . . . . . . . . . . . 104
3.3 Thermal Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.3.1 Heat Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
3.3.2 Heat Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
3.3.3 Measurement of Specific Heats . . . . . . . . . . . . . . . . . . . . 118
3.3.4 External Determination of Heat Transfer . . . . . . . . . . . . . 122
3.3.5 Internal Determination of Heat Transfer . . . . . . . . . . . . . 123
3.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
4 The First Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4.2 Internal Energy and the Energy Equation . . . . . . . . . . . . . . . . . . . 130
4.2.1 The Equivalence of Work and Heat . . . . . . . . . . . . . . . . 131
4.2.2 Steady State Problems . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.2.3 Change of State Problems . . . . . . . . . . . . . . . . . . . . . . . . 136
4.3 The Energetic and Enthalpic Equations of State . . . . . . . . . . . . . . 138
4.3.1 Liquids and Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
4.3.2 Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
4.3.3 Liquid–Vapor Equilibrium . . . . . . . . . . . . . . . . . . . . . . . 150
4.4 The Open System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
4.4.1 Steady Flow Problems . . . . . . . . . . . . . . . . . . . . . . . . . . 155
4.4.2 Steady Flow Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.4.3 Variable Mass Systems . . . . . . . . . . . . . . . . . . . . . . . . . . 167
4.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5 The Second Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5.2 Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
5.3 The Entropic Equation of State . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Contents xiii
5.3.1 Liquids and Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

5.3.2 Gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
5.3.3 Liquid–Vapor Equilibrium . . . . . . . . . . . . . . . . . . . . . . . 191
5.4 The Irreversibility Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
5.4.1 Entropy Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
5.4.2 Calculation of Entropy Generation . . . . . . . . . . . . . . . . . 195
5.4.3 Open Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
5.5 Heating and Power Bounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
5.5.1 Constraints on Heat Transfer . . . . . . . . . . . . . . . . . . . . . 204
5.5.2 Constraints on Work . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
5.5.3 The Carnot Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
5.5.4 Refrigerators and Heat Pumps . . . . . . . . . . . . . . . . . . . . . 216
5.5.5 Thermodynamic Efficiency . . . . . . . . . . . . . . . . . . . . . . . 219
5.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
6 Power and Refrigeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
6.2 Vapor Power Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
6.2.1 The Newcomen Engine . . . . . . . . . . . . . . . . . . . . . . . . . 231
6.2.2 Watt’s Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
6.2.3 The Rankine Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
6.3 Air Standard Power Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
6.3.1 The Otto Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
6.3.2 The Diesel Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
6.3.3 The Brayton Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
6.4 Refrigeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
6.5 Vapor Refrigeration Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
6.5.1 The Ideal Vapor Compression Cycle . . . . . . . . . . . . . . . . 268
6.5.2 Subcooling and Superheating . . . . . . . . . . . . . . . . . . . . . 269
6.5.3 Compressor Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
6.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Appendix A: Thermodynamic Properties: English Units . . . . . . . . . . . . . 279

Appendix B: Thermodynamic Properties: SI Units . . . . . . . . . . . . . . . . . 297
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Chapter 1
Measurement and Properties of Matter
1.1 Introduction
Dynamics, as you have learned in previous courses, is the study of how particles and
rigid bodies move under the action of the forces and moments that are exerted on
them. Thermodynamics is the theory that integrates dynamics with the concept of
temperature. Dynamics is a fundamental part of physics because of Newton’s second
law, which describes how the momentum of a fixed quantity of matter evolves in
time. Thermodynamics is fundamental because of its first law, which describes how
the energy of a fixed quantity of matter evolves in time, and its second law, which
describes the asymmetry of this evolution. The energy equation is needed whenever
we wish to describe thermodynamic processes; motions that result from an induced
temperature change, or temperature changes that occur as a consequence of motion.
The science of thermodynamics is applied by engineers to study and understand
the energy exchanges that take place between bodies as they undergo thermodynamic
processes, in order to be able to design devices that operate efficiently and econom-
ically. Examples of such devices are, compressors, turbines, engines of all types, air
conditioners, and power generating stations.
Thermodynamics can be developed on either one of two levels; microscopic or
macroscopic. On the microscopic level we think of matter as a collection of an
enormous number of discrete particles that move according to equations of motion,
and consider their aggregate behavior based on an additional statistical postulate.
This viewpoint leads to statistical thermodynamics, in which temperature and other
macroscopic variables are described in statistical terms and depends on the character-
istics of the particles and their interactions. This viewpoint is the more fundamental
of the two, however, it is also conceptually and mathematically more difficult. On
the macroscopic level we think of matter as a continuous distribution that moves in
accordance with an equation of motion, and postulate an additional energy evolution
equation. This viewpoint leads to continuum thermodynamics, in which temper-
ature and other variables describe the state of the continuum, and are related by
© Springer Nature Switzerland AG 2020 1

A. M. Whitman, Thermodynamics: Basic Principles and Engineering Applications,
Mechanical Engineering Series, https://doi.org/10.1007/978-3-030-25221-2_1
2 1 Measurement and Properties of Matter
experimentally determined equations of state. Usually this viewpoint, which is the

one we will study, is presented without reference to the underlying atomic structure
of matter; however, I will make these connections where I feel that it will help your
understanding.
Since I will need to introduce you to new state properties during the course of our
study, some of which will be unfamiliar, I shall begin with an introductory discussion
of some familiar properties, and the measurement processes by which we quantify
them. Much of this material you have studied before; my approach here is meant to
reinforce what you have already learned well, and help clarify what you have not.
1.2 Dimensions and Units
The concepts of length (space interval) and duration (time interval) are familiar to
us from mechanics as well as experience. A collection of lengths or durations can be
put in sequential order by directly comparing them pairwise, and observing which
is longer. A length however, can not be directly compared with a duration, they
are elements of different dimensions. By dimension we simply mean an orderable
collection of things. When needed, we denote the dimension of a quantity by a symbol
enclosed in square braces, [L] for length and [t] for duration.
When we place a certain length, or duration, into sequence, we say that we have
measured it. For example, if we have already established a scale of lengths
L0 < L1 < L2 < L3 < L4
where L0 denotes an absence of length, and the Lj represent physical lengths, deter-
mining that a new length L satisfies the inequalities L2 < L < L3 , constitutes a mea-
surement of L. We say that L is greater than L2 and less than L3 . The closer the
lengths L2 and L3 are to each other, the higher is the precision of the measurement.
The elements of the scale, the standards of comparison, are called standards.
When the dimension to be measured and the elements of the scale are not adjacent,
it is necessary to make indirect comparisons via a transportable intermediate device.
In doing this we are using the property of transitivity. For example, if L is a length
to be measured, Ls is a scale element length, and LT is the device length
[L = LT and LT = Ls ] =⇒ L = Ls (1.1)
which is valid provided that the dimension is unchanged when the device is trans-
ported, LT = LT 1 = LT 2 . This fundamental assumption is adopted for all the indirect
comparisons we make in classical physics. The device used for [L] is a rule, and for
[t] is a clock.
The measurement of lengths and durations are quantified by using the additivity
property that each of these dimensions possesses (this is an important point because
not all dimensions are additive). This means that when either a length, L, or a duration
1.2 Dimensions and Units 3
t, is subdivided into N parts, with Lj and tj the length and duration of the j th part,
the whole is equal to the sum of its parts
N
N

L= Lj t= tj (1.2)
j=1 j=1
These equations tell us that lengths and durations can be added and equated. Indeed
these mathematical operations correspond to physical acts. Likewise, it makes no
sense to either add or equate terms corresponding to different dimensions. Using
these equations, the process of making a length or duration measurement can be
described as follows:
1. Select a single standard length, Ls or duration ts for comparison purposes.
2. Identify L0 , t0 with 0, consistent with the addition rule.
3. By direct comparison determine how many standard quantities must be combined
to produce the required length or elapsed time.
If the process produces n repetitions of Ls with an additional fraction f remaining,
Eq. (1.2) gives
n
L= Ls + fLs = (n + f )Ls (1.3)
j=1
and we say that the length L is n + f times the standard length Ls .1 The additivity
property has allowed us to create an infinite number of standards from one. Thus, a
single standard is sufficient to specify the entire scale, and every length is defined
as a (numerical) multiple of the standard. The selection of standard quantities is
completely arbitrary, and over the years many different scales of measurement have
been devised and used. What is important for scientists, engineers, manufacturers,
traders, etc., is the ability to relate them.
Time Scales
Historically the universally accepted standard for duration measurement is the time
interval in which the earth rotates once about its axis, the day (d). This quantity is
too long to conveniently express many commonly occurring time intervals, so it is
divided into 86,400 equal seconds (s). This relationship is, from Eq. (1.2),
1 d = 86400 s (1.4)
where we have used t = ts = 1 d, and tj = 1 s. We make these replacements because

the names and abbreviations have more meaning to us than the abstract symbols,
1 Actually f is not determined exactly, rather we establish that f− < f < f+ , and we say that L is
greater than n + f− and less than n + f+ times Ls . The difference f+ − f− = 2p is an indication of
the precision of the measurement, and is usually included compactly in reporting the result of the
measurement by writing [(n + f ) ± p] Ls , where f = (f+ + f− )/2.
however, because they are symbol replacements, we can treat the names as we do
any other algebraic quantity. For example, we can divide each side of Eq. (1.4) by 1
d to obtain s
1 = 86400 (1.5)
d
Notice also that the unit names are crucial to the meaning of the equation, which
would be mathematically incorrect without them
1 = 86400
Equations (1.4) and (1.5) are called conversion equations because they allow us to
convert from one unit to another. For example, 0.1 day is easily converted into 8640 s
by multiplying it by 1, in the form of Eq. (1.5). Other scales and their unit names
are introduced, by means of appropriate conversion equations so that we can always
express our measurements in terms of numerical factors that are close to unity. In the
present case of duration, the month and year are introduced because of their intimate
connection with life on earth.
The entire spectrum of units is called a system of time units. One of them, the
second, is regarded as the basic or base unit and the rest as subsidiary units. There
is nothing fundamental2 in this, it merely reflects the fact that a measurement can
be expressed in terms of any unit by means of the appropriate conversion equation.
Moreover, as is clear from this example, the base unit and the standard need not be
the same.
Length Scales
Today the universal standard for [L] is the meter (m)3 which is also used as a base
unit for a spectrum of subsidiary length scales that are all related by powers of ten,
and whose names have Greek or Latin prefixes that indicate their power
1 cm = 10−2 m
1 km = 103 m
This decimal system was introduced in France at the time of its revolution, and,
because of the ease with which we can multiply and divide by ten, has replaced
non-decimal systems almost everywhere in the world. The USA is a singular, but
important exception. Here the base unit is the foot (ft), the conversions to other units
in the spectrum are numbers with many factors, and the unit names are historical
2 The second is chosen because it can be defined with greater precision than other standards in
this atomic age. Its definition is 1 s = 9,192,631,770 periods of the radiation corresponding to the
transition between the two hyperfine levels of the ground state of the cesium 133 atom.
3 The modern definition of the meter is the distance light travels in 1/299,792,458 s in a vacuum.
1 in = 1/12 ft
1 mi = 5280 ft
There is nothing to recommend this system, except that it has been in use for centuries,
and, although the decimal system is judged better by virtue of its simplicity, its
acceptance in the USA will take a long time due to economic, infrastructure, and
psychological factors. Meanwhile, it is important to be able to convert a measurement
using one of these base units to the other. This conversion equation was defined in
1959 by an international agreement to be4
1 ft = 0.3048 m (1.6)
exactly.
Dimensional Equations
In order to avoid errors in writing equations involving dimensions, a few simple rules
and conventions must be observed
1. The numerical value of a dimensional quantity must never be written without
including the associated unit.
2. An equation must be dimensionally homogeneous; only like dimensions may be
added or equated. However, dimensions may be multiplied and divided at will,
and the resulting composite dimensions are the product and quotient, respectively.
The quotient of two like dimensions is dimensionless.
3. An equation should be unitwise homogeneous. Using this convention, the terms
of an equation will add as if they were dimensionless, and the numerical value on
each side of an equality will be the same.
The following example illustrates these rules and conventions.
Example 1.1 If L1 , and L2 are lengths, and t1 , and t2 are times,
?
L31 = t1 L2
is not a valid equation, because it is not dimensionally homogeneous. The

composite dimensions [L3 ] = [L] [L] [L], and [Lt] = [L] [t] are not the same.
t1 2 ?
L21 + L = L1 L2
t2 2
4 Historically theyard, equal to 3 ft, was a standard length maintained by the British government in
the Tower of London, and the meter was a standard length maintained by the French government
at its Bureau of Standards in Sèvres. This is no longer the case as noted in the previous footnote.
is valid because it is dimensionally homogeneous. The composite dimension

of the second term [ tL2 / t] = [L2 ] is the same as the dimension of the other
terms. Note that dimensions can be manipulated like ordinary algebraic quan-
tities.
L1 ? t1
=
L2 t2
is valid because it is dimensionally homogeneous. The composite dimension

of each term [ L/ L] = [1] and [ t/ t] = [1] is dimensionless.
The equation
1m + 1ft = 1.3048m
is dimensionally correct, and true; but inconvenient to use, because the numer-
ical values of the terms do not add. We convert all terms to the same units by
putting Eq. (1.6) into the form (1.5). This produces the dimensionless (but not
unitless) equation
1 = 0.3048 m/ft
which we use as we indicated previously
1m + 1 f t × 0.3048m/ f t = 1.3048m
When it is in unitwise homogeneous form the terms add numerically and

there is numerical equality on each side of the equation. Note that an equation
that is homogeneous unitwise is necessarily dimensionally correct.
One way to ensure unitwise homogeneity is to express all quantities in an equation

in terms of base units. If you are ever in doubt, use this option.
1.2.1 Fundamental and Derived Dimensions
Other dimensions are also familiar from experience. One that was already of some
practical importance to our ancestors is area, [A]. Area, like length and duration,
satisfies an order relation (a larger area overlaps a smaller), and is additive. Conse-
quently we can choose a standard area, say the area of a certain square, with which
to compare all others, and using it, we can make measurements, as we did previously
with length and duration.
Now it was clear to astute traders, and proved by early geometers, that the owner of
a square of land twice as large on each side as his neighbor, was four times wealthier
(had four times the area for planting or grazing, see Fig. 1.1).
This fact can be expressed more generally in mathematical terms as a ratio,
A/A1 = (L/L1 )2
Fig. 1.1 An example of the relation between the areas of two squares and the length of their sides
On cross-multiplying, this takes the form
A/L2 = A1 /L21 ≡ kS
or written as a dimensional equation involving only one square
A = kS L2 (1.7)
where A is the area of the square whose sides have length L, and kS is a dimensional
constant having dimensions, [A/L2 ]. There are two possible interpretations that we
can make of Eq. (1.7)
1. Area and length are independent dimensions, like length and duration. Then the
specific value of kS depends on the units used for the area and length measure-
ments. For example, if an arbitrary size square is measured in acres, and the length
of one of its sides in ft, the values, when substituted into Eq. (1.7) give
kS = 2.296 · 10−5 acres/ft2 (1.8)
Equation (1.7) may subsequently be used with this value of kS to determine the
area, in acres, of other squares whose side length is known in ft.
A = (2.296 · 10−5 acres/ft 2 )L2
The units must be written explicitly, as we have done, once the specific value of
kS has been substituted. With Eq. (1.8) for kS , and the value of L expressed in
units of ft, the value of A produced by the calculation is acres. Of course other
area and length units may be used, with an appropriate value of kS .
2. The constant kS is taken as a pure (dimensionless) number. Accordingly the dimen-

sion [A] is equivalent to [L2 ]. We say that [A] is a secondary or derived dimension,
whereas [L] is a fundamental dimension. With the usual choice, kS = 1, Eq. (1.7)
becomes
A = L2 (1.9)
This is dimensionally consistent since both sides of the equation have dimension
[L2 ]. The natural units are an appropriate composite of the base units of the
fundamental dimensions, and in this case are either m2 or ft2 . Of course, we can
still express area in acres, which in this interpretation is simply another subsidiary
unit. We do this by using conversion equations as we illustrated in Example 1.1. To
get a conversion equation between acres and ft2 we substitute the measurements
described above into Eq. (1.9) and find
2.296 · 10−5 acres = 1 ft2 (1.10)
The numerical value calculated for area is the same no matter which of these inter-
pretations is used. You can see this immediately by setting kS to one in Eq. (1.8)
and observing that the result is the same conversion factor as you get on dividing
Eq. (1.10) by ft2 . Moreover, whether a dimension is fundamental or derived is a matter
of definition. In our example here of area, we could have said that area is fundamen-
tal and length a derived dimension, rather than the other way around. Furthermore,
dimensions that are regarded as fundamental now may subsequently be found to be
related. In such a case we can either continue to regard them as fundamental (as in
interpretation 1), or adopt one of them as derived (as in interpretation 2). Finally,
I note that in either interpretation, the existence of a relation among two or more
dimensions, like Eq. (1.7) here, implies that one of them need not be measured, but
can be calculated in terms of measurements of the others.
Systems of Measures
In order to avoid having many independent dimensions, and associated dimensional
constants cluttering our equations, interpretation 2 is adopted here. Length is retained
as a fundamental dimension while area is taken to be a derived dimension. There are
two spectra of subsidiary area and length units, systems of measures, in common use
in the world today. Each one is based on a single standard, the meter. As we indicated
previously, most countries use a decimal system in which the meter itself is the base
unit, the natural unit for area in this system is m2 . The system used in the USA has
the foot as base unit with the natural area unit ft2 .
Minimization of the number of fundamental dimensions is the usual preference,
but as we shall see it is not always convenient, and thus not always done. In any event,
the decision is simply a matter of choosing the number of fundamental dimensions
we wish to have in a system that comprises many interrelated dimensional quantities,
and is a part of establishing a consistent system of units. As we discuss additional
dimensional quantities, we will add to the lists of both fundamental and derived
dimensions.
Location
0 0.5 1.5 2 in Distance

0 1.27 3.81 5.08 cm from O'
Distance
0 1 1.5 in Distance
2.54 3.81 cm from O

0
Fig. 1.2 The location of a point on a line is a differentially additive concept. It can be quantified,
as distance from another point, in an infinite number of ways
1.2.2 Absolute and Relative Quantities
The additivity property, Eq. (1.2), was a crucial element in establishing numerical
scales for the measurement of length and duration. Now there are other measurable
quantities of interest that do not themselves have that property, whereas differences in
them do. Such differentially additive concepts can be quantified in an infinite number
of ways. This is a new feature that requires some explanation.
Let the location of a point A on a line be denoted by ζA , and a point B by ζB , see
Fig. 1.2. When A is to the right of B, ζA > ζB , so that location is a dimension and is
measurable in the general sense of Sect. 1.2. Location does not possess the addition
property, so we cannot define a scale for it that depends on a single standard, as we
did previously for length, duration, and area; however, the difference between two
locations can be expressed in terms of the space interval (length, see Sect. 1.2.1)
between them
ζA − ζB = xA/B > 0
This means that location has the same dimension as length, [L], and that the difference
between two locations is additive. Thus we can quantify location as the distance from
an arbitrary origin. We call distance a relative quantity because it has an arbitrary
origin. Quantities that do not have an arbitrary origin are called absolute. Each specific
distance, for example, the distance from O, denoted xo , is a different realization of
location, but they all have the same difference
ζA − ζB = xA/B = xA/O − xB/O = xA/O − xB/O
This is true generally, the difference between two values of every realization of a
differentially additive quantity is absolute, and leads to the conversion equation
xo = xo + xO /O (1.11)

Table 1.1 Approximate mileage between major eastern seaboard cities

Boston New York Philadelphia Washington
Boston − 200 300 400
New York 200 − 100 200
Philadelphia 300 100 − 100
Washington 400 200 100 −
The distinction between absolute and relative quantities is important when we need
to form ratios. Referring to Table 1.1, we see that the answer to a question about
distance ratios, how many times farther is Boston than New York, cannot be given
absolutely. A Philadelphian would answer three times, and a Washingtonian would
answer two times. That is why we call distance a relative quantity. On the other hand,
both the Philadelphian and the Washingtonian would agree that it is twice as far from
Boston to New York as it is from New York to Philadelphia; a question about length
ratios does have an absolute (observer independent) answer. That is why we call
length an absolute quantity.
Time is, for all practical purposes, a relative quantity; the origins of various time
realizations (calendar time), are placed arbitrarily. However, duration, a time interval,
is an absolute quantity. Duration shares with time the same relation that length shares
with distance. In Fig. 1.3 we illustrate the relation between several calendar times in
use today, along with an absolute time that is based on a current physical theory of
the origin of the universe. This shows that one can define and use realizations with
arbitrary zeros whether or not an absolute quantity exists. Here the relation among
the times is, from Fig. 1.3
Big Creation of Birth of Flight of

Bang ? Adam Jesus Muhammad
Epoch
Absolute
0 Time ?
Jewish
0 3761 4383 yr Time
Christian Calendar
0 622 yr Time Time
Islamic
0 yr Time
ABSENCE OF TIME ?
Fig. 1.3 Several realizations of calendar time, a relative quantity, compared with a hypothetical
absolute time
ta = toj + tj = toj + 3761yr + tc = toj + 4383yr + ti
Even though toj (the absolute time of Adam’s creation) is not known precisely,5 we
can still relate the calendar times to each other. For example, tj = 3761yr + tc , from
the second equation above. Of course, one cannot answer all questions absolutely
in terms of a relative quantity, therefore if an absolute quantity exists it is worth
knowing it.
As we introduce more dimensions into our scheme we will need to consider and
use both absolute and relative quantities in our equations.
1.3 Properties of Matter
Matter we recognize as existing in various gaseous, liquid, or solid forms. In any of

these physical and chemical forms, we characterize it by those properties that we
can measure. Consequently those properties are, like length and time, dimensional
quantities.
1.3.1 Volume
The simplest property of a fixed collection of matter, which we call a closed system
or body is its volume, V . If the matter is chemically identifiable, we may alternatively
refer to it as a substance. Like area, volume is taken to be a derived dimension, with
the volume of a cube of side, L, given by
V = L3
and as a result the natural units for volume are either m3 or ft3 . There is a spectrum
of subsidiary units, for example, a gallon (gal) or a liter (l).
1 gal = 0.1337 ft3 (1.12)

1 l = 10−3 m3 (1.13)
which are included with lengths and areas in their respective system of measures.
We can use the additive property of volume to write for a body that is subdivided
into N parts, each with volume Vj ,
5 Current cosmological measurements give the age of the universe as 13.799 billion±21 million
years.
N

V = Vj (1.14)
j=1
We call volume an extensive, meaning additive, property. Since we consider matter

to be continuously distributed, we can imagine an infinite sequence of ever smaller
subdivisions, and in the limit we write an integral in place of the sum

V = dV (1.15)
V
Here V, denotes the set of points enclosed by the surface of the body. If the matter
undergoes a process, its volume may change in time. We write V (t) when we want
to emphasize this possibility.
As you have already learned in Calculus, Eq. (1.15) for volume has a counterpart
for the other measures we have discussed. For example, the area enclosed by a plane
curve, and the arc length of a curve segment are, respectively,

A= dA L= dL (1.16)
A L
1.3.2 Weight and Mass
Experience tells us that everybody is attracted to the earth. We call the intensity of
this attraction the weight of the body. We recognize that weight depends on both
the earth and the body, whereas we think of mass as the source, within the body, of
the attraction. Mass is therefore intrinsic to the body while weight is a result of the
mutual interaction of the body and the earth.
The mass of a body remains constant during all thermodynamic processes of
interest, however, it is possible to consider a fixed volume in space, and ask how
much mass it contains. If the matter is in motion, the amount of mass contained in
this open system may change, M (t). Indeed we will find it profitable to consider open
systems later in this course.
In the event that the system contains more than one substance or more than one
physical phase of a substance, the mass of each constituent is counted as a property.
When chemical reactions or phase changes occur the sum of the constituent masses
remains constant; this is the principle of conservation of mass.
Both mass and weight are measurable quantities; they can each be ordered by
using a balance (a heavier mass descends and a lighter mass rises). Moreover they
are additive (mass is an extensive property); on subdividing the body as we did with
volume
N N
M = Mj W = Wj (1.17)
j=1 j=1
1.3 Properties of Matter 13
or on subdividing indefinitely

M = dM W = dW
M M
where M is the set of mass points comprising the body. Moreover, it follows from the
fact that both mass and weight are measured by a balance, along with their additivity,
Eq. (1.17), that weight and mass are proportional (e.g. if the mass is doubled so is
the weight at any place on the earth). However, the constant of proportionality is
different at different places on the earth.6 Taking all this into consideration, we write
(refer to Sect. 1.2.1)
W = ke fM = Ke M = f Ŵ (1.18)
where ke is a dimensional constant that characterizes the earth, with dimensions

[W]/[M], f is a pure number that represents the effect of changing place on the body
weight, and Ŵ is a nominal weight which will differ from the actual weight of the
body, W , as f varies.
As we discussed previously (Sect. 1.2.1), there are two alternatives for ke . Unfor-
tunately both alternatives are used, and this causes some confusion about the relation
between weight and mass. In either event mass, [M], because it is intrinsic to matter,
is chosen as the fundamental dimension. As with length there are two base units in
common use, the kilogram (kg), and the pound (lb). The kg is the unit name for the
internationally accepted standard mass, the kilogram, which is the mass of a certain
platinum–iridium cylinder kept at Saint-Cloud, France.7 The lb was defined, by the
same international agreement that defined the ft, to be
1 lb = 0.45359237 kg (1.19)
exactly.
Systems of Weights and Measures
In order to promote commerce, sovereign governments regulate weights (and cur-
rency) in addition to measures of length, area, and capacity (volume). This means
adopting a choice for ke as well as choosing standards, base units, and accepted
subsidiary units with their associated conversion factors (see Sect. 1.2.1). Of the two
practical systems of weights and measures in use in the world today, both choose ke
to be the pure number unity. Thus in each system the dimension weight, [W], is the
same as mass, [M], indeed Eq. (1.18) gives in this case, Ŵ = M , indicating that in
6 Weight is not transitive (see Sect. 1.2) so we can not use a balance to compare weights at different
locations.
7 At the time this was written, November 2018, a new definition of the kilogram, based on specifying
the value of Plank’s constant as h = 6.62607015 × 10−34 kg m2 /s, is being proposed to the General
Conference on Weights and Measures (CGPM), which is an international body tasked with acting
on matters related to measurement science and measurement standards on behalf of its member
nations. When this definition is accepted there will be no need to keep a standard mass.
these systems there is no distinction between weight and mass. In practical metric
systems the natural weight unit is the kg and in the USCS (United States Custom-
ary System) it is the lb; a person when asked will give his weight in kilos (kg) or
pounds (lb), respectively. These practical systems, which are perfectly satisfactory
for our daily use involving weighing with balances, are inconvenient for scientific
application.
In science, it is useful to distinguish between the dimensions weight, [W], and
mass, [M]. Historically this was done by defining the standard weight, either the
kilogram force (kgf) or the pound force (lbf) as the attraction of a kilogram mass
(kg) or pound mass (lbm ≡ lb) at a standard place on the earth. Therefore in these
systems ke has dimensions [W]/[M], and its value is either 1 kgf/kg or 1 lbf/lbm by
definition. In other words, at the standard place, where f = 1, the numerical value
of the standard weight is identical with the standard mass. However at other places
on the earth, the standard mass will not be attracted with the same intensity. We will
discuss this further in the section on force.
1.3.3 Density and Specific Volume
From experience we know that equal volumes of different substances have unequal
masses (weights). A property that quantifies this characteristic is the mass intensity
or density, ρV , defined as the amount of mass that occupies a given region of space
M
ρV = (1.20)
V
Density is therefore a derived dimension, [M/L3 ], with natural units kg/m3 , and
lbm/ft3 . Values for density are determined by making measurements of both mass
and volume. According to Eq. (1.20), if a system is subdivided into parts, each part
has density
ρj = Mj /Vj (1.21)
which, in general, will vary from part to part. In the case of continuous distributions
of matter, this process can be continued indefinitely, and produces a finite density
for every point, x,8
ρ(x) = lim Mj /Vj
Vj →x
Making use of Eqs. (1.17), and (1.21) in (1.20) we see that the system density is
a weighted average (not the sum) of the individual densities
N N
1 1
ρV = Mj = ρj Vj
V j=1 V j=1
8 Bold face letters designate vectors. Here the symbol denotes the vector from an origin to the point
in question.
A property with this quality is called intensive. Multiplying both sides by V gives
an equation for the calculation of mass from a knowledge of the density
N

M = ρj Vj
j=1
which for continuous distributions of matter is an integral

M = ρ(x, t) d V (1.22)
V
When we want to emphasize the local nature of intensive properties, we include their
arguments, ρ(x, t). In this way we display their dependence on location as well as
time, a characteristic that distinguishes them from extensive properties that depend
on time only.
The specific volume, v(x, t), is defined as the inverse of density
v(x, t) = 1/ρ(x, t)
and is used much more frequently in Thermodynamics than the density itself. In
virtually all of the applications of interest to us, the density is a constant, ρ, and
therefore we have
1 V
v= = (1.23)
ρ M
Notice that in this special case ρV = ρ. Equation (1.23) is extremely simple, yet it is
of major importance in what we will do later.
1.3.4 Velocity and Acceleration
When a particle, P, moves in a straight line, its distance, xP/O , from an origin, O,
changes in time. The average velocity of P with respect to O is a vector whose
direction is the direction of the line, and whose measure number is the intensity of
the change (the length of the move divided by its duration)
xP/O (t2 ) − xP/O (t1 )

vP/O = vP/O i = i
t2 − t1
Here i is a unit vector parallel to the line; when vP/O is positive it is the direction
of vP/O , and when vP/O is negative the direction of vP/O is −i. According to this
definition, the average velocity is a derived dimension equal to [L/t]. This average
velocity, like distance from a specific origin (see Sect. 1.2.2), is a realization of a
differentially additive quantity. Indeed we can define many velocities, each relative
to its own origin and direction. The result, as you have learned in mechanics, is that
the ratio of the magnitude of two velocities has no absolute meaning. The question,
how many times faster than the bus is the limo traveling, will elicit different answers
from a pedestrian and the bus driver.
The average velocity depends on the duration measurement, but as the duration
becomes arbitrarily small, the ratio approaches a finite limit that we call the instan-
taneous velocity (we drop the subscripts in the interest of simplicity, but you should
keep in mind that velocity refers to a particular particle, origin, and fixed vector)
dx
v = vi = i
dt
Of course, we can only measure average velocities by simultaneously measuring the
length traveled and duration. For short durations these measurements approximate
the instantaneous velocity, which we usually refer to simply as the velocity.
When the particle moves along a curve in space the instantaneous velocity is
a vector that can be described in terms of the distances x, y, z in three mutually
perpendicular directions, i, j, k, from an origin
dx dx dy dz
v= = i+ j+ k
dt dt dt dt
As you learned in dynamics, this is a vector tangent to the curve, whose magnitude
2 2 2
dx dy dz
|v| = + +
dt dt dt
gives an indication of how fast the particle is moving at any instant, relative to an
observer at rest at the origin, and for whom the unit vectors, i, j, k, are constant.
In a system of N particles, each one has a velocity, say vj (t). When we wish
to give only a global description of the motion, we use a weighted (by the masses)
average
N
1
vc = mj vj
M j=1
which we call the velocity of the mass center.

For a continuous distribution of matter, there is a velocity at each point, v(x, t),
and the velocity of the mass center is given by the integral

1
vc (t) = ρ(x, t)v(x, t) d V (1.24)
M V
When a particle moves, its velocity also changes in time. The acceleration is
defined by the change in velocity similarly to the way that velocity is defined by the
change in position
dv d 2x
a= = 2
dt dt
From this you can see that acceleration is a derived dimension equal to [L/t2 ]. More-
over it is a relative quantity like distance and velocity, since no unambiguous zero
can be defined.
For continuous distributions of matter, the mass center has an acceleration that is
defined analogously to the velocity, as given by Eq. (1.24).
1.3.5 Force
The concept of force, like mass length and time, comes to us both from mechanics
and experience. We regard a force as a push or pull that is exerted on a body at its
surface, and we recognize that when a force acts on a body, the body will move
or deform (or both). The direction of a force, as well as its intensity (magnitude),
is important, so we need to describe forces by vectors. If we are given two forces
with the same direction, we can decide which is larger by reversing one of them and
applying them both to a body at rest (the body will begin to move in the direction of
the larger force). Accordingly force is a dimensional quantity, [F]. Moreover force
is additive and an absolute quantity; its absence is unambiguous, and defined to be
zero.
Since weight can cause a body to fall or deform, we regard weight as the magnitude
of a force, even though it does not act at the surface. It is a body force in contrast with
the contact forces discussed previously. This conception of weight was quantified
by Isaac Newton who initiated classical physics with his two laws that describe the
motion of mass points relative to an inertial observer
N
mj
G = kG m rj (1.25)
j=1
rj3
d
F + G = kA mv = kA ma (1.26)
dt
As usual kG and kA are dimensional constants that are needed to make each of these
equations [the law of gravitation, Eq. (1.25), and the law of motion, Eq. (1.26)]
dimensionally consistent. The quantities F and G are the resultants of the contact
and body forces, respectively, that act on the mass point, whose mass is m and inertial
acceleration is a. The quantity rj that appears in Eq. (1.25) is the distance from m to
another mass point, mj that attracts it.
From Eq. (1.25) we can represent what was earlier called weight with the attractive
force exerted by a spherical, symmetric earth on a small (compared to the size of the
earth) body, of mass M , near its surface. Thus
me
Ge = −kG M ke = −W ke (1.27)
R2e
where ke is a unit vector in the direction of the local vertical, me is the mass of the
earth, and Re is the distance from the mass M to the center of the earth. Comparing
this with Eq. (1.18), W = Ke M , shows that
me
Ke = kG
R2e
According to this understanding, the weight defined in Sect. 1.3.2 should be called
the earth weight and denoted We = Ke M , because if the mass were transported to
the surface of the moon its moon weight would be different, Wm = Km M , where
mm
Km = kG
R2m
Indeed Eq. (1.27) produces an earth weight that varies slightly depending on location,
so that on comparing with Eq. (1.18), W = ke fM
ke = kG me /R∗2
e f = (R∗e /Re )2 (1.28)
in which R∗e is the radius of the earth at the standard place described in Sect. 1.3.2,
Systems of Weights and Measures.
Scientific Systems of Units
In order to use Newton’s laws in equation form, we need to make choices for the
dimensional constants appearing in them. This means choosing a system of units that
includes the dimension force. In making this choice, the constant kA is taken to be
the pure number unity for the reasons discussed with interpretation 2. of Sect. 1.2.1.
However, this means that if we used the practical system of weights and measures,
in which mass [M] and weight [W] (the same as force [F]) are identical dimensions,
Eq. (1.26) would require length [L] to be a derived dimension equivalent to [t2 ].
Although there is nothing wrong with this in principle, we are so used to thinking
about length and time as independent dimensions, that this alternative is unacceptable
to virtually everyone. Therefore force (consequently weight also) is taken to be a
derived dimension, defined through Eq. (1.26) as [ML/t2 ]. The natural units are
either kg m/s2 in the SI (Systèm International d’Unités), or lbm ft/s2 in the English
System.9 These have the names Newton (N) and poundal, respectively,
9 We have used the name English System here because it is widely used. However, the only major
country in which it is used is the USA. All other English speaking countries use the SI for scientific
and industrial work.
1 N = 1 kg m/s2
1 poundal = 1 lbm ft/s2
The Newton is the force unit used in the SI system, however, the poundal is seldom
used. The subsidiary unit lbf, which we will describe further on, is most common in
the English System.
The constant kG is not taken to be the pure number unity, for the same reason
that we do not use the practical measurement systems; if we did, Eq. (1.28) would
require mass [M] to be a derived dimension equivalent to [L3 /t2 ]. Consequently mass
[M], length [L], and time [t] are independent dimensions, Ke has dimensions [L/t2 ],
and its value can be obtained by measurements on freely falling bodies (for which
weight is the only force acting). It is found using Eqs. (1.26) and (1.28), that these
accelerate toward the earth at a constant rate dependent on the place but not on their
mass
dv
Ke = − M
M =Ma ≡M g
dt
a relation first noted by Galileo. In terms of this “acceleration due to gravity”, the
formula for (earth) weight, W = Ke M , is
W = Mg (1.29)
Today a standard g is defined, ge , with the values 9.807 m/s2 or 32.174 ft/s2 . Com-
paring with Eq. (1.18), W = ke fM , we find
ke = ge f = g/ge
These define the standard place and the subsidiary force units (see Sect. 1.3.2,
Systems of Weights and Measures)
1 kgf = 9.807 kg m/s2

1 lbf = 32.17 lbm ft/s2 (1.30)
The kilogram force, kgf, is not used in the SI system. The pound force, lbf, is used
in the English System and care must be taken to distinguish it from the pound, lb,
that designates both weight and mass in the practical system, USCS.
Dynamicists use a subsidiary mass unit, the slug, defined such that
1 slug = 32.17 lbm
and they take this to be the base unit for mass. Their reason for doing so is that then
the lbf becomes the natural force unit, similar to the Newton in the SI system
1 lbf = 1 slug ft/s2 (1.31)

Although the slug is commonly used in courses on dynamics and fluid mechanics,
the lbm is used exclusively in thermodynamics and heat transfer. This is unfortunate
and causes confusion among students, however, if you always distinguish between
lbf and lbm, and if you remember that the lbf is a subsidiary unit that often must
be removed from an equation when making it homogeneous unitwise [using either
Eq. (1.30) in thermodynamics and heat transfer or Eq. (1.31) in dynamics and fluid
mechanics], you should be able to keep things straight.
You may still encounter difficulty in other peoples use of lb. In such cases note the
dimension of the quantity and translate to lbf or lbm as appropriate. For example a
150 lb person becomes a 150 lbf person because weight is a force [F], while a density
of 62.4 lb/ft3 becomes 62.4 lbm/ft3 (or 1.94 slug/ft3 ) because density has dimension
[M/L3 ]. This also works for weights given in kilos, which become kgf and should be
converted to Newtons in the SI system.
Force Measurement
According to our choice of kA in Eq. (1.26), force has become a derived dimension
whose measurement results from knowing what acceleration it imparts to a specified
amount of mass. However, it is inconvenient to measure forces in this way. Moreover,
as we have discussed, the standard weight cannot be moved unchanged from the
standard place. Therefore, the easiest way to measure the magnitude of a force, F,
is to measure the result of the action of the force on a standard body. Such a body,
which deflects under the action of a force, is called a force gauge, or gauge.
The basis for this indirect measurement is the relation between the gauge deflec-
tion, xg , and the magnitude of the force acting on the gauge, F
xg = xg (F)
however we do not measure xg explicitly, rather we introduce a linear force measure

by means of the relation
Fx = C1 + C2 xg (F)
This measure will be absolute if and only if we can specify the deflection that corre-
sponds to no applied force. This we can do in terms of the free length, x0 . If we also
measure the length xG when the gauge is deformed by a weight, WG , the relations
0 = C1 + C2 x0
WG = C1 + C2 xG
allow us to solve for the constants C1 and C2 to obtain
xg − x0
Fx = WG (1.32)
xG − x0
There is no need to measure xg and convert to force units by means of this equation;
we simply inscribe the force values directly on the gauge. Of course the measured
force will reflect the true value in all cases only if the gauge material obeys a linear
(Hooke’s) law
xg − x0 = F/ks
In that case Eq. (1.32) produces Fx = F for all values of F, meaning that the gauge
correctly measures the force.
Example 1.2 A force gauge is constructed using a spring that deflects accord-
ing to the relation
F
xg = x0 + 1 + εF 2
ks
Determine the relation between the force, Fx , indicated by the gauge, and the
force, F, exerted on the gauge, in terms of the calibration weight, WG . This
is called a calibration curve. For the values ε = 0.01 lbf−2 , and WG = 10 lbf,
what force does the gauge indicate when 5 lbf is exerted on it?
Solution. The gauge deflection under the calibration weight is found by sub-
stituting the appropriate values into the given deflection formula
WG
xG = x0 + 1 + εWG2
ks
From this we find that

xg − x0 F 1 + εF 2
=
xG − x0 WG 1 + εWG2
and on substituting into Eq. (1.32), Fx = WG (xg − x0 )/(xG − x0 ),

1 + εF 2
Fx = F
1 + εWG2
In the accompanying figure we plot, in dimensionless form, the difference

between this function and the straight line Fx = F which a perfect gauge
would indicate. In this plot we used the given values, so
εWG2 = (0.01 lbf −2 )(100 lbf 2 ) = 1
Then, if 5 lbf is exerted

1 + 25 lbf 2 /100 lbf 2
Fx = 5 lbf
1+1
or
Fx = 3.125 lbf
The largest error in the range
0 ≤ F ≤ WG
√
occurs at F = WG / 3. A gauge with smaller values of ε and WG will have a
smaller maximum error.
1.3.6 Impulse and Momentum
The law of motion for a particle, or the mass center of a rigid body, Eq. (1.26), can
be integrated in time
t2 t2 t2
d
F dt + G dt = mv dt
t1 t1 t1 dt
The integral on the right can be evaluated in terms of a property we call the momentum
P = mv
so we can write
IF + IG = P2 − P1 (1.33)
The two integrals on the left, which have been denoted by a subscripted letter I,
are the impulse of the resultant contact and gravitational force, respectively. The
equation is known as the Impulse–Momentum form of the equation of motion. It
states that the impulse of the resultant force acting on a body during a specified time
interval is equal to the change in the momentum of the body.
The momentum of a body is a property with dimension [ML/t]. It is a relative
quantity since velocity itself is relative; in addition it is extensive. The impulse of a
changing force, F, t2
IF = F dt
t1
also has dimension [ML/t], however, in general, it is neither a property of the force
nor of the body; it depends on the motion of the body. We call the impulse of a force,
and other quantities which, like it, are defined as integrals over time that depend
on the motion, interactions. Interactions are important to us as engineers since they
are the quantities we must provide in order to bring about a desired change in some
property.
1.3.7 Work and Energy
Another integral of the equation of motion of a particle, or mass center of a rigid

body, can be obtained by first dot multiplying by the velocity
t2 t2 t2
d
F · v dt + G · v dt = v· mv dt
t1 t1 t1 dt
On making use of a well known relation from differential calculus
d d 1 2
v· mv = mv
dt dt 2
the integral on the right can be evaluated in terms of a property we call the kinetic
energy
1
K = mv2
2
and we write the equation as
WF + WG = K2 − K1 (1.34)
The two integrals on the left, which have been denoted by a subscripted letter W ,
are the work done by the resultant contact and gravitational force, respectively. The
equation is known as the Work–Energy form of the equation of motion. It states that
the work done by the resultant force during a specified time interval is equal to the
change in kinetic energy of the particle.
The kinetic energy is an extensive, relative, property of dimension [ML2 /t2 ]. The
natural energy unit, kg m2 /s2 , in the SI system is called a Joule (J). The natural unit
in the English System, lbm ft2 /s2 , has no name. Results are most often expressed in
terms of the subsidiary unit, ft lbf.
The work done by a force F that moves with velocity v
t2
WF = F · v dt
t1
is an interaction with the same dimension as kinetic energy.
1.3.8 Pressure
When two bodies are in contact across a finite area, or if one body is imagined
separated into two parts across an area, the force exerted by one body (part) on the
other is actually distributed across the entire area. This is important because applying
the same force across different areas produces different reactions of a body. As an
example think about pressing the eraser end or the point end of a pencil against
your thumb with the same force. As a consequence, we are lead to consider a force
intensity, or traction, which quantifies this characteristic. Notice that this is similar
to the way we were lead to introduce density as a mass intensity
F
τ nA =
A
From this definition we see that traction is a vector, and that it has dimensions of
force per unit area, [F/A], or [M/Lt2 ]. Moreover, both its magnitude and direction
depend on the direction of the normal to the surface on which it acts, n (see Fig. 1.4).
Similarly to what we did in the case of density, we can subdivide the surface into
Fig. 1.4 The traction on an element of surface has a direction different from the normal direction,
and at a point, P, is different for two surface elements with different normal directions
parts. Then the traction on each part is
τ nj = Fj /Aj
and subdividing indefinitely we obtain a traction for each point and direction
τ n (x) = lim Fj /Aj

Aj →x
When we know the traction at all points, we can calculate the contact force acting
on a body bounded by a surface A, by integration

F(t) = τ n (x, t) dA (1.35)
A
Here n is the outward pointing normal to the surface at the point x.

In general the traction at a point has a direction different from the normal to the
surface element, Fig. 1.4, and it is resolved into normal and tangential components
called the normal stress and shear stress, respectively10
τ n (x, t) = σn (x, t)n + τn1 (x, t)t1 + τn2 (x, t)t2 (1.36)
For fluids (gases and liquids) at rest the shear stresses are zero, a fact that allows
matter in these phases to adapt to the shape of their containers. Static equilibrium
then requires that the normal stress at a point be independent of the orientation of the
surface, moreover, experience shows that this normal stress is always compressive.
We call the positive, isotropic normal stress the pressure. Thus the traction at a point
in a fluid at rest is simply
τ n (x) = −p(x)n (1.37)
for any direction n. Using the definition of τ n it is clear that pressure is defined by
force and area,
p(x) = lim F/A
A→x
and is an absolute quantity. It is a local variable, and is, like density, intensive.
The customary units for expressing pressure are the Pascal (Pa), which is the
natural unit in the SI
1 Pa = 1 N/m2 = 1 kg/m s2
or the pound per square inch (psi) in the English System,
1 psi = 1 lbf/in2 = 1 lbf/in2 · 32.174 lbm ft/lbf s2 · 144 in2 /ft2 = 4633 lbm/ft s2 .
10 The traction at a point is dependent on the direction of n, while the stress tensor, T, is not. That
is what makes the stress tensor a physically important quantity, and is why you spent a lot of time
learning about it in your course on strength of materials.
Fig. 1.5 On the left is a pressure gauge which operates on the elasticity of a spring, and on the
right is a gauge that operates on the elasticity of a gas
Pressure Measurement
We can measure the average pressure exerted by a fluid on the wall of its container
by replacing a small part of the wall with a movable section, or piston, attached to
a force gauge, Fig. 1.5. The spring will deflect until static equilibrium is achieved
among the forces acting on the piston, and the measured pressure, pg , is then given
by
pg = Fx /A
where Fx is the force indicated by the gauge, and A is the piston area. This area
should be as small as possible so that the measured value is as close as possible to the
pressure at the point. Note that pg will be the same as p = F/A only if Fx is the same
as F. However, if the air in the gauge exerts a pressure pa , then static equilibrium of
the moving section (ignoring friction) shows that
pg = p − pa
This means that gauge pressure is a relative quantity (see Sect. 1.2.2) even though
pressure is an absolute quantity, as indicated in Fig. 1.6. When the measured gauge
pressure is positive, the actual pressure is higher than the atmospheric pressure, and
when it is negative the pressure is lower than atmospheric. In the later case, we say
that there is a partial vacuum.
Fig. 1.6 Illustrating the Perfect Atmospheric

relative nature of gauge Vacuum Pressure
pressure. Compare this with
Figs. 1.2 and 1.3 which show 0 101.32 kPa Absolute
relative distance and time 0 14.696 psi Pressure
0 kPa Gauge
0 psi Pressure
ABSENCE OF PRESSURE
As an alternative to the spring, we could use the elasticity of gases to make the
measurement, Fig. 1.5. In this case the piston will move, as before, in response to the
unbalanced forces; however, as a result of the decreasing volume of the gas in the
measuring tube, the pressure there will increase. Correspondingly as the volume of
the measured gas increases its pressure will decrease. When the pressures on either
side of the piston are equal, it will cease moving and the pressure can be determined
from its location
h = h(p)
This discussion shows that the act of measurement changes the pressure that we
wish to measure. The volume of the instrument should be small compared with the
volume of the measured substance in order that the measured pressure be nearly the
pressure in the undisturbed state. Introducing a linear measure as we did with force
(see Sect. 1.3.5, Force Measurement)
pg = C1 + C2 h(p)
If we set the measured pressure to zero when h is zero (this is arbitrary), and note the
location hG corresponding to a standard pressure, obtained from a standard weight
WG /A, we can evaluate the constants to produce the expression
WG h
pg =
AhG
For measuring low pressures we do not use the tube device as shown in Fig. 1.5,
but rather a slight modification, called a manometer and shown in Fig. 1.7. Here when
the pressure in the tube increases, it displaces a volume of liquid until the weight
of the unbalanced part equals the applied force, at which point static equilibrium
is achieved. This means that, with constant liquid density, WG = ρghG A, and the
Fig. 1.7 A manometer on the left, and a barometer on the right

measured pressure is
pg = ρgh = ρg(ho − hi ) (1.38)
In this equation the gauge pressure is zero when the unbalanced height of the liquid
is zero. If the atmosphere exerts a nonzero pressure, pa , then static equilibrium of
the liquid manometer column requires that
ρgh = p − pa
and therefore this gauge does not measure the pressure, but like the spring gauge
discussed before, the difference between the pressure and the atmospheric pressure,
pg = p − pa . Indeed all pressure gauges work this way.
Example 1.3 A mercury manometer connected to a tank of air reads ho of 2.0

in and hi of −2.0 in (refer to Fig. 1.7). What is the gauge pressure in psi?
Solution.
Since the density of mercury is 849 lbm/ft3 , substitution into Eq. (1.38) pro-
duces
1
pg = 849 lbm/ft3 32.17ft/s2 [2 − (−2)] in = 9104 lbm/ft s2
12 in/ft
Expressing this in the units of psi
9104 lbm/ft s2
pg = = 1.96 psi
4633 lbm/psi ft s2
As you can see here manometers are used to measure small gauge pressures.
Atmospheric Pressure
One way to obtain an absolute measure of pressure, is to create a condition of no
pressure corresponding to the scale zero. This can be accomplished approximately
by means of a device called a barometer, and also shown in Fig. 1.7. The device
is constructed by completely filling a tube with a liquid and inverting it with the
open end submerged. The space at the top of the tube is not a perfect vacuum (zero
pressure) because some atoms of the manometric substance exist there. Nevertheless,
the pressure is very low and is considered to be zero for practical purposes. The
atmospheric pressure is then equal to the weight of the liquid column divided by the
tube area
pa = W/A = ρgh (1.39)
In order to obtain the absolute pressure of any system we must make two mea-
surements, the atmospheric pressure, pa , and the gauge pressure pg . Then
p = pa + pg (1.40)
As I mentioned previously, this is true for any type of gauge pressure measurement.
Note that although the difference between two absolute pressures is the same as the
difference between the corresponding two gauge pressures
p1 − p2 = pg1 − pg2
(provided the atmospheric pressure is the same for the two measurements), the ratio
of the two absolute pressures is not the same as the ratio of the two corresponding
gauge pressure measurements
p1 /p2 = (pa + pg1 )/(pa + pg2 ) = pg1 /pg2
This behavior is characteristic of all relative quantity realizations (see Sect. 1.2.2).
Example 1.4 A horizontal cylinder with a sliding piston 6 inches in diameter,

contains a gas whose pressure is indicated by a gauge as 35 psi. The piston is
maintained in equilibrium by an external force, and the atmospheric pressure
is 15 psi. What is the value of the equilibrating force? How many times greater
is the cylinder pressure than the atmospheric pressure?
Solution.
Since the piston is in static equilibrium, the forces acting on it, specifically
in its direction of possible motion, must sum to zero. These forces include the
force exerted by the gas, pA (A is the area of the piston), the force exerted by
the atmosphere, pa A, and the external force that maintains the equilibrium, F
(the free body diagram is shown on the left). In equation form this is
pA − pa A − F = 0
or F = (p − pa )A. Now the cylinder pressure is given by Eq. (1.40), so the

force is F = pg A. Substituting the piston area (πD2 /4), 28.27 in2 , and gauge
pressure into this expression gives the force as 990 lbf. The cylinder pressure
is not equal to the value read on the gauge; according to Eq. (1.40) it is 50 psi
in the present case. This is 50/15, or 3.333 times greater than the atmospheric
pressure. Notice that this is not the same as the ratio of the cylinder gauge
pressure to the atmospheric pressure, 35/15, or 2.333.
In this course whenever I refer to a pressure without specifically stating that it is

measured by a gauge, you may assume that I am referring to the absolute pressure.
The atmospheric pressure results from the weight of the column of air above the
place where the measurement is made. Therefore, assuming a constant density for
air, it is
pa = Wa /A = ρa gha
where ρa is the air density and ha is the height of the atmosphere. We can measure
object heights by applying this formula at two vertical locations
H = ha1 − ha2 = (pa1 − pa2 )/(ρa g)
The meaning of the symbols appearing in this formula are illustrated in Fig. 1.8.
Example 1.5 The barometric pressure at the base of a building is 100.7 kPa
and on the roof it is 97.36 kPa. If the density of air is 1.18 kg/m3 , what is the
height of the building?
Solution. Using the formula we get
(100.7 − 97.36) · 103 kg/m s2

H= = 289 m
1.18 kg/m3 · 9.807 m/s2
for the height of the building.
Fig. 1.8 Altitudes can be measured this way assuming the atmosphere has a constant density, and
therefore a finite height. This is represented by the dot-dashed line in the figure
Fig. 1.9 The force exerted

on a particle by the right wall
of its container. The shaded
area is the impulse, and is
equal to the lighter shaded
area whose height (dotted
line) is the average force
Kinetic Model for Pressure

If a substance is regarded as composed of particles that can move about freely, then
pressure can be related to this underlying motion. In the simplest possible model we
consider a cubical volume, V , containing Np particles each of mass, mp . The number
of particles, Np , is a pure number, not a dimensional quantity, so that both the total
mass
M = Np mp
and the particle mass, mp , have dimension [M]. We assume that 1/3 of the particles
are traveling parallel to each of the three spatial directions, that they all have the
same speed, c, that they do not collide with one another, and that collisions with
the walls are perfectly elastic (meaning that on impact with the walls, the particle
velocity direction is reversed and its magnitude remains the same).
The Impulse—Momentum form of Newton’s second law for a single particle
striking a wall is11
tf
Ip = Fp dt = mp (vf − vi ) = mp (−c − (+c)) = −2mp c
ti
and the time for a round trip is

tr = 2V 1/3 /c
According to this, the force exerted by the wall on the particle, Fp (t), which is
sketched in Fig. 1.9 is a very complicated function of time, but is periodic with
period tr . Consequently the average force can be written simply as
ta +tr
1 Ip mp c2
{Fp } = Fp (t) dt = = − 1/3
tr ta tr V
11 we write only the component of the vector equation for the direction normal to the right vertical
wall, Sect. 1.3.6. The positive direction is to the right and is outward pointing.
The total force exerted by the wall is obtained by summing over all the particles that
hit the face
Np 2 Np mp c2 1
F= {Fp } = −
3 3 2 V 1/3
and the pressure is the average force per unit area of the face V 2/3
2 Mc2 /2
p= (1.41)
3 V
From this equation we see that the pressure is just a fraction of the kinetic energy
per unit volume, an intensive variable.
1.3.9 Heating, Hotness, and Temperature
Our senses indicate to us the concept of heating, Q̇, from which we infer the property
of hotness of a body, φ (hotter bodies heat cooler ones).
When two bodies, H and L, are placed in contact so that one can heat the other,
we say that the bodies are in thermal contact, and that the region between them is a
conductor. Conversely if the region is an insulator the contact is adiabatic and there
is no heating. The heating is high when the hotness difference, φH − φL , and contact
area, A, are large, it becomes less as they become smaller, and it is zero when the
two bodies are equally hot. The state of equal hotness is called thermal equilibrium.
All this can be summarized by
Q̇L = ht A(φH − φL ) = −Q̇H (1.42)
in which Q̇H and Q̇L are the heating of H and L, respectively. The heating, also called
the rate of heat transfer, is an additive dimensional quantity with dimension [Q/t]. The
standard Q is, like the standard duration, ephemeral and likewise is used by making
indirect comparisons. It can be defined as the heat transfer which causes a specified
mass of a standard substance to melt. The degree of hotness is also a dimensional
quantity, [T], while ht , of dimension [Q/A T t], characterizes the heating process and
has a value that depends on the material of the conductor. The new dimensions, [Q]
and [T], are regarded here as fundamental, but [Q] ultimately (in chapter 4) will be
viewed as a derived dimension (see Sect. 1.2.1).
We can define many standard degrees of hotness, such as the melting of ice, φi ,
and the boiling of water, φs , which satisfy an order relation, φi < φs , based on hotter
bodies heating colder ones. However, hotness does not have the usual additivity
property; if a body is subdivided arbitrarily, the sum of the hotness of each part is
not equal to the hotness of the complete body. Therefore, a scale consisting of a
single standard cannot be defined as we did previously in the case of length, mass,
etc. (recall the discussion in Sect. 1.2). A simple way of putting all this is that given
two bodies, we cannot say when one is twice as hot as the other. In other words,
although we know how to obtain a length of 2 m from the standard meter, a mass
of 2 kg from the standard kilogram, and a time of 2 s from the standard second, we
do not know how to obtain a hotness that is twice that of melting ice, 2φi , from the
standard melting ice hotness, φi .
According to Eq. (1.42), hotness is a differentially additive quantity like location
and epoch (see Sect. 1.2.2) so we can quantify it, as we quantified location by
distance, and epoch by calendar time in Sect. 1.2.2, by first arbitrarily assigning a zero
hotness, and subsequently a standard hotness interval. The resulting relative quantity
we call empirical temperature (or just temperature). There are two realizations of
temperature in use today. These are the Fahrenheit temperature, Tf , for which the
hotness of a melting ice–salt mixture is zero, and the Celsius temperature, Tc , for
which the hotness of pure melting ice is zero. The standard interval for both is taken
as the hotness difference between boiling water and melting ice with 180 units, called
degrees Fahrenheit, ◦ F, assigned for Tf , and 100 units, called degrees Celsius, ◦ C,
assigned for Tc . The general relation that characterizes the realizations of all relative
quantities (see Sect. 1.2.2), here applied to temperature
φH − φL = THf − TLf = THc − TLc
when specialized to the ice and steam points produces the conversion factor between
the two units, 1 ◦ C = 1.8 ◦ F. In addition this relation shows that the dimension of
temperature is [T], the same as hotness. Historically, relative temperatures were used
because early workers were unaware of the existence of an absolute cold. However,
with the development of thermometry and thermodynamics, it became apparent that
this was a useful and valid concept, so an absolute temperature, T , whose zero coin-
cides with an absence of temperature, is now also used by scientists and engineers.
The relations among these temperatures are illustrated in Fig. 1.10 (see also
Sect. 1.2.2, and Figs. 1.2, 1.3, 1.6). The base unit used for Fahrenheit temperature is
the degree Fahrenheit, that for Celsius temperature is the degree Celsius, while two
Absolute Freezing Ice Ice Steam

Cold Salt Mixture Point Point
Hotness
0 273.15 373.15 K Absolute

0 459.67 491.67 671.67 R Temperature
Fahrenheit
0 32 212 oF Temperature Empirical
Temperature
Celsius
0 100 oC
Temperature
ABSENCE OF TEMPERATURE
Fig. 1.10 A comparison among the various temperatures we have defined. The two large numbers
on each scale denote the standards used to construct it
base units are used for absolute temperature, the kelvin, K, (SI) and the rankine, R,
(English). Given any one of the three temperatures, the other two can be calculated
using ⎫
T = Tc + 273.15K 1K = 1 ◦ C ⎬
T = Tf + 459.67R 1R = 1 ◦ F (1.43)
⎭
Tc = Tf − 32 ◦ F 1 ◦ C = 1.8 ◦ F
which all follow from Fig. 1.10. Different symbols are needed to designate each of
these temperature realizations, because they correspond to different zero locations.
Example 1.6 At a Celsius temperature of 25 ◦ C, what are the corresponding

Fahrenheit and absolute (on both kelvin and rankine scales) temperatures?
Solution. All required results are obtained from Eq. (1.43)
T = 25 ◦ C · 1K/1 ◦ C + 273.15K = 298.15K
To calculate the Fahrenheit temperature we use the last equation, but first solve
explicitly for Tf . You must always solve equations explicitly in order to make
your calculations — including the units — simpler and more direct.
Tf = Tc + 32 ◦ F
Tf = 25 ◦ C · 1.8 ◦ F/1 ◦ C + 32 ◦ F = 77 ◦ F
Notice how the conversion factor is used here so that we add terms that are
homogeneous unitwise (see Sect. 1.2). The absolute temperature in rankine can
either be calculated from the kelvin value, 298.15 K, found above, by using
the conversion factor 1K = 1.8R, or the middle equation of Eq. (1.43)
T = 77 ◦ F · 1R/1 ◦ F + 459.67R = 536.67R
Note that all three equations listed in Eq. (1.43) are not independent.
Substituting Tc = xc ◦ C and Tf = xf ◦ F, along with the conversion factor between

Celsius and Fahrenheit degrees into the last of Eq. (1.43), and canceling the units
produces
xf = 1.8xc + 32 (1.44)
This dimensionless equation is sometimes used to convert from one relative temper-
ature value to the other, and to answer questions about their relationship. Equation
(1.44) is particularly easy to use by noting that 1.8 = 9/5.
Fig. 1.11 Two types of thermometer; on the left a liquid in glass thermometer, and on the right a
constant volume gas thermometer
Example 1.7 At what body temperature is its Fahrenheit temperature value

twice the corresponding Celsius value?
Solution. Here we want the temperature for which xf = 2xc . Using this in
Eq. (1.44) gives
xf = 1.8xf /2 + 32
Solving for xf gives the Fahrenheit temperature as 320 ◦ F.
Thermometry
Equation (1.42) does not provide a convenient basis for measuring temperature dif-
ferences, just as Newton’s law was found to be inconvenient for measuring force (see
Sect 1.3.5, Force Measurement). Moreover, experience shows that the properties of
material bodies, including hotness, are interrelated; they all vary with changes in any
one of them.12 In this event we resort to the use of standard bodies with which to
make indirect measurements13 as we did previously with force and pressure.
In Fig. 1.11 we show two typical temperature gauges or thermometers. Each of
these operates on the principle that when the hotness of the thermometer increases,
so does some other property of the thermometric substance.
Liquid in Glass Thermometer When the thermometer bulb is maintained in diather-
mal contact with a hotter system whose temperature is to be determined, the hotness
difference causes heating of the thermometer. As the hotness of the thermometric
liquid increases so does its volume, and it expands to occupy a greater length of the
tube. Concurrently the hotness of the measured system decreases. When the hotness
12 We will study this fully in the next chapter.

13 In connection with temperature measurement, the transitivity assumption required for all indirect
comparisons [see Eq. (1.1)] is given a special name, the zeroth law of thermodynamics.
of each are the same, heating, as well as all other change, ceases. The common tem-
perature can then be determined in terms of the length of the tube filled with liquid,
z(T ). We note that as with pressure measurement, the act of measuring temperature
changes the temperature to be measured. Furthermore, if the temperature surround-
ing the thermometer bulb is not uniform, the thermometer will indicate an average
temperature. For both reasons a thermometer should be small, in some sense, relative
to the system whose temperature it is used to measure. We will be able to quantify
this notion later on when we study heating in more detail.
Rather than use the length z, which depends on the construction of the thermome-
ter, as a measure of temperature it is more physically meaningful to use temperature
itself. This is done, as we did before with force and pressure, by adopting a linear
relation between the measured temperature, Tzg (g here denotes, as it did with pres-
sure, a relative quantity), and length, z(T ), and defining values of temperature at two
standard levels in order to determine the constants in
Tzg = C1 + C2 z(T )
For a particular thermometer (construction and liquid), we measure the column length
at the hotness of melting ice, φi , and boiling water, φs , at a standard atmospheric
pressure, call the two temperatures Tig and Tsg , respectively (the g subscript indicates
relative temperatures),
z = zi when Tzg = Tig z = zs when Tzg = Tsg
and evaluate the constants in terms of these quantities [see also Eq. (1.32)]

z − zi
Tzg = Tig + (Tsg − Tig ) (1.45)
zs − zi
The Celsius and Fahrenheit temperatures are obtained by assigning specific values
to the ice point, as I have described (see Fig. 1.10)

z − zi ◦ z − zi
Tzc = (Tsc − Tic ) Tzf = 3 2 F + (Tsf − Tif ) (1.46)
zs − zi zs − zi
From Eq. (1.46) we obtain the last of Eq. (1.43) for the pair Tzc , Tzf . Note that it is
not necessary to measure the column length and use the formula to calculate Tzc and
Tzf , because these numbers are inscribed directly on the gauge, see Fig. 1.11. It is
clear from Eq. (1.45) that, except at the standards, thermometers containing liquids
that expand differently will produce different temperature values when they are in
contact with a body of specified hotness.
Since both the Celsius and Fahrenheit temperatures are relative (zero on either
is unrelated to the absence of temperature), when using them we cannot say abso-
lutely how many times hotter boiling water is than melting ice. It is infinitely hotter
according to the Celsius temperature (100/0), but only about six and a half times
hotter (212/32) by the Fahrenheit temperature.
Although the liquid in glass thermometer is commonly used, and familiar to us in

our daily existence, temperature measurements based on it suffer from two defects
• Only relative temperatures can be measured, no information can be obtained about
absolute temperature.
• With the exception of the two standards, Tig and Tsg , each thermometric liquid
produces a (slightly) different value of temperature when placed in thermal contact
with a body of specified hotness.
These defects can both be removed by using an alternative gauge for temperature
measurement.
Constant Volume Gas Thermometer Although we cannot produce a condition
corresponding to absolute cold, for use as a standard, an absolute temperature can be
devised by making use of a condition that corresponds to an absence of temperature.
This is done by using a gas as the thermometric substance and measuring its absolute
pressure while keeping its volume fixed. This last condition is necessary because we
know from our previous discussion of pressure measurement that the gas pressure
changes as a result of volume changes. Since the absolute pressure in a gas is always
greater than or equal to zero, so we can make temperature measured in this way also
greater than or equal to zero, with zero corresponding to an absence of temperature.
The beauty of the gas thermometer is that with it, we can measure both relative and
absolute temperatures, and thus find the relation between them.
A constant volume gas thermometer can be constructed as shown on the right in
Fig. 1.11. The volume of the gas (in the bulb, tubing, and space above the mercury
column) in this device is maintained constant by adjusting the height of the mano-
metric fluid reservoir, and the absolute pressure, p(T ), is measured. Assuming, as
usual, a linear relation, the relative temperatures, in terms of Tig and Tsg defined
previously [see Eq. (1.45)], for which the pressures are pi and ps , respectively, are

p − pi ◦ p − pi
Tpc = (Tsc − Tic ) Tpf = 32 F + (Tsf − Tif )
ps − pi ps − pi
while the absolute temperature is simply

Tp = Toc + Tpc = Tof + Tpf
We can show from these that the last of Eqs. (1.43) is true for the pair Tpc , Tpf ,
and Eq. (1.44) is true as well. The relation between the Fahrenheit and Celsius
temperatures is the same for different thermometer types. But we can discover even
more. Since Tp = 0 when p = 0 we find using two of these equations
Toc = (Tsc − Tic )pi /(ps − pi ) (1.47)
and therefore
Tp = Toc (p/pi )
Consequently the absolute temperature is proportional to the measured pressure ratio,

p/pi , while Toc , the absolute temperature, or hotness, of the ice point depends on the
unit size, Tsc − Tic , and the measured pressure ratio ps /pi .
Fig. 1.12 The ratio of measured pressures ps /pi in the extrapolated limit of no gas in the ther-
mometer is independent of the thermometric gas. This is true independent of the hotness of the
body being measured, as is shown on the right
Although these measurements vary slightly from one gas to another, the variation
becomes smaller as the mass of gas in the thermometer is decreased. The results of
such experiments, which are illustrated in Fig. 1.12, show that to a high degree of
accuracy the limiting process produces a single value
lim (ps /pi ) = 1.3661

pi →0
and a unique (gas independent) value for absolute temperature
T = lim Tp = Ti lim (p/pi ) (1.48)

pi →0 pi →0
where Ti = limpi →0 Toc is obtained from Eq. (1.47) by substituting the limiting value
for ps /pi . Historically the Kelvin scale was defined with Ts − Ti = 100K between
the ice and steam points. Therefore 1 K = 1 ◦ C and
Ti = 100K/0.3661 = 273.15K
as given in Eq. (1.43). The Rankine scale has 180 R between the ice and steam points
so that its ice point value is 180 R/0.3661 = 491.67 R and 1 R = 1 ◦ F.
The unique value given by Eq. (1.48) is called the ideal gas temperature.14 It
depends only on the hotness of the body in equilibrium with the thermometer, not
on the specific thermometric gas. Accordingly, we calibrate all other thermometer
readings by comparing them to the corresponding ideal gas values. Henceforth when
14 In1967 the ideal gas temperature was defined in terms of the hotness of the triple point of water
(the condition at which the solid, liquid, and vapor phases are simultaneously in equilibrium),
TT = 273.16 K. This number was chosen to maintain other characteristics of the scale close to their
previously established values, Ti = 273.15K and Ts − Ti = 100 K.
we refer to temperature we mean the ideal gas temperature, and presume that a value
measured by another means has been calibrated appropriately.
The difference between absolute and relative values is important, as we have seen
before, when we form temperature ratios
T2 /T1 = (Tc2 + 273.15K)/(Tc1 + 273.15K) = Tc2 /Tc1
but unimportant when we form temperature differences
T2 − T1 = Tc2 − Tc1 = Tf 2 − Tf 1 (1.49)
We find from Eq. (1.48) that boiling water is 36.61% hotter than melting ice, and
also what hotness is twice Ti (the one for which limpi →0 p/pi = 2).
The ideal gas temperature is not in itself a theoretically satisfactory measure of
hotness because it depends on the concept of an ideal gas. However, in chapter 5,
based on thermodynamic arguments, we will discover another, completely substance
independent, measure for hotness. When applied to substances that are ideal gases,
this thermodynamic temperature is found to be identical with the ideal gas tempera-
ture, a coincidence that justifies the use of the latter as a standard.
Thermocouples and Resistance Thermometers There are a number of devices
that make use of a change in some electrical property of a body with a change in its
hotness. These devices are useful because they are small (and so do not much affect
the temperature they are intended to measure), and because their output is readily
adapted to the input of a computer, thus allowing measurements to be recorded and
manipulated without human intervention.
The basis for operation of a resistance thermometer is the relation between the
electrical resistance, R = R(T ), of a small element of the material and the ideal gas
temperature. Modern systems do not use this relation to form a gauge temperature,
TRg as in Eq. (1.45), rather they invert it by computerized data reduction circuitry
using a polynomial expression for T (R)
T = a0 + a1 R + a2 R2 + · · · + an Rn
The coefficients aj and the power n depend on the thermometric material, and are
supplied by the manufacturer; however, the user need know none of these details
since the output reading is T .
The device used most often for temperature measurement in engineering labora-
tories is called a thermocouple. It consists of two wires made of different metals and
joined together at one end. Between the two free ends, there is a voltage, V = V (T ),
that varies with the absolute temperature of the junction. When two thermocouples,
one of which is maintained at Ti , are connected together, the voltage difference is
V = V − Vi = V (T ) − V (Ti )
Computerized data reduction circuitry is used to invert this relation for T , using the
measured V value, known Ti , and polynomial approximations for V (T ) and its
inverse T (V ). Finally T is supplied directly to the user.
Kinetic Model for Temperature
Although hotness is a notion which is foreign to mechanics, and kinetic models
attempt to describe properties based on a mechanical conception of matter, it is
nevertheless possible to assign a kinetic interpretation to it. This possibility derives
from the relation between temperature and pressure that constitutes the ideal gas
temperature measurement; for using Eq. (1.41), the kinetic expression of pressure,
in the definition of the ideal gas temperature, Eq. (1.48), produces
T = Ti lim (c/ci )2 = Ti (c/ci )2

Np →0
This equation implies, in the limit, either
T ∝ c2 or T ∝ mp c2 /2
We choose to use the second of these because energy is a more fundamental quantity
than velocity squared. We will see in Chap. 2 that this choice is consistent with devel-
opments on the macroscopic scale. Then forming an equation from this proportion
1 2 1
T = kT mp c2 = mp c2 (1.50)
2 3kB 2
Note that if we take the dimensional constant kT to be a pure number, then temperature
becomes a derived dimension equivalent to [E]. However if we take kT to have
dimensions [t2 T/ML2 ] = [T/FL] = [T/E], then temperature remains a fundamental
dimension. The second alternative is adopted in both the SI and the English System,
and for the same reason that mass is retained as a fundamental dimension. The
dimensional constant kB , which is more useful than kT , is called the Boltzmann
constant in honor of Ludwig Boltzmann, one of the developers of the statistical
viewpoint of thermodynamics. It has the value
kB = 1.380649 × 10−23 J/K
Equation (1.50) indicates the nature of the condition corresponding to zero abso-
lute temperature. It results when all motion of the particles cease. In that event there
is zero pressure as well, since the particles do not impact the walls.
1.3.10 Coefficient of Thermal Expansion
The size of the volume change for a given temperature change, with pressure remain-
ing constant, differs for different substances. We call it the coefficient of thermal
expansion, and define it as a fractional change, so that we will be able to compare
the coefficient of thermal expansion of two substances that may have quite different
densities
1 ∂V 1 ∂v 1 ∂ρ
α= = =− (1.51)
V ∂T p v ∂T p ρ ∂T p
For most substances and over most temperature ranges, α is positive, namely when
temperature increases substances usually expand.15
The coefficient of thermal expansion is the property that is the basis of the liquid
in glass thermometer. A substance for which this property behaves as
α = C/v (1.52)
has a linear volume–temperature relation, found from integrating Eq. (1.51), with
Eq. (1.52) for α
v − vi = C(T − Ti )
Here vi is the specific volume at the ice point. The length of a cylindrical column of
such a liquid is therefore a linear function of temperature, and
z − zi T − Ti
=
zs − zi Ts − Ti
Substituting this into Eq. (1.45), shows that a thermometer filled with such a ther-
mometric liquid would always agree with the ideal gas temperature,
Tzg − Tig = T − Ti
or specifically Tzc = Tc . In actuality no substance behaves as Eq. (1.52) requires, and

no liquid thermometer agrees completely with the ideal gas thermometer.
1.3.11 Compressibility
The size of the volume change for a given pressure change, with temperature remain-
ing constant, also differs for different substances. That is to say, it is also a property.
Although this is the isothermal compressibility of the substance, we call it here just
15 The most notable exception to this rule is water between 0 ◦ C and 4 ◦ C at atmospheric pressure.
the compressibility, and define it as the fractional volume change for a unit change
in pressure at constant temperature,

1 ∂V 1 ∂v 1 ∂ρ
β=− =− = >0 (1.53)
V ∂p T v ∂p T ρ ∂p T
Whenever we increase the pressure on a substance its volume becomes smaller (recall
the discussion of pressure measurement in connection with Fig. 1.5), so we include the
negative sign in the definition in order to make the compressibility a positive number.
Indeed, the compressibility must be positive; no material that expands as pressure
increases exists. Clearly from the definition, compressibility is an intensive property.
The compressibility of liquids are much smaller than gases; for a given pressure
change, they incur a much smaller fractional volume change. For that reason, liquids
are said to be relatively incompressible.
Of course, in order to measure the compressibility we must use differences rather
than the derivative
1 V
β=−
V p
Here the difference V2 − V1 > 0 should be small enough that we are satisfied with
the precision
1 V2 − V1 1 V2 − V1
− <β<−
V2 p2 − p1 V1 p2 − p1
Compressibility is a measure of the elasticity of a substance. For linear elastic solids

β = 1/G, where G is the bulk elastic modulus you learned about in strength of
materials.
1.4 Exercises
Section 1.3.8
1.1 A manometer filled with mercury is connected to a tank, and has an ho of 2.3
in and an hi of −2.3 in. What is the gauge pressure in the tank? (2.26 psi)
1.2 A manometer filled with water is connected to a tank, and has an ho of 5 in and
an hi of −5 in. What is the gauge pressure in the tank?
1.3 A pressure gauge reads 23.7 psi when a barometer stands at 29.85 in of mercury.
What is the absolute pressure? (38.3 psi)
1.4 A mercury manometer which is connected to a vacuum chamber has an ho of

−4.25 in and an hi of 4.25 in. If the atmospheric pressure is 14.60 psi, what is the
absolute pressure in the chamber?
1.4 Exercises 43
1.5 A barometer reads 30.01 in mercury at the base of a mountain, and 19.55 in
mercury at its peak. How high is the mountain (use the value 0.0737 lbm/ft3 for the
density of air)? (10020 ft)
1.6 A skin diver exploring coral reefs at a depth of 60 ft experiences what hydrostatic
(gauge) pressure (use 64 lbm/ft3 for the density of seawater)?
1.7 A pilot takes off from an airport where the barometric pressure is 29.92 in
mercury. At his destination, the reading on his altimeter is 1500 ft. The tower reports
the local pressure is 28.86 in mercury. When he adjusts his altimeter to reflect this
value, what will it read (use the value 0.0737 lbm/ft3 for the density of air)? (484 ft)
1.8 The crushing strength of marble is about 15000 psi. What is the highest mountain
of this material that can exist?
1.9 An 8 in diameter horizontal cylinder contains a gas enclosed by a sliding piston.

What force must be exerted on the piston in order to maintain a gauge pressure in the
cylinder of 15.3 psi. Neglect friction and use 14.7 psi for the barometric pressure.
(769 lbf)
1.10 A vertical cylinder is capped by a 0.5 in thick steel sliding piston. If the atmo-
spheric pressure is 10 psi, what is the pressure in the contained gas?
1.11 A vertical cylinder is capped by a 1 cm thick steel sliding piston. If the atmo-
spheric pressure is 100 kPa, what is the pressure in the contained gas? (100.8 kPa)
1.12 A vertical cylinder is capped by a 6 in diameter 0.75 in thick aluminum sliding

piston. If the atmospheric pressure is 11.3 psi, what is the pressure in the contained
gas? What is the weight of the piston?
1.13 A manometer filled with mercury is connected to a tank, and has an ho of 2

cm and an hi of −2 cm. What is the gauge pressure in the tank? (5.32 kPa)
1.14 What is the atmospheric pressure at the top of the Washington Monument
(elevation 169 m) when it is 101 kPa at the base (use the value 1.18 kg/m3 for the
density of air)?
1.15 A 0.04 m diameter vertical cylinder is capped by a sliding piston. The gas
pressure inside is 300 kPa, and the barometric pressure is 101.3 kPa. Neglecting
friction what is the mass of the piston? What does a pressure gauge give as the
cylinder pressure? (25.5 kg, 198.7 kPa)
1.16 A piston that can slide horizontally in a cylinder is held in equilibrium against a
250 kPa gauge pressure by a 12.7 N force. What is the piston area? If the atmospheric
pressure is 100 kPa, what is the cylinder pressure?
Section 1.3.10
1.17 Calculate the Celsius, Fahrenheit, and absolute (Rankine) temperatures of a
substance at 318K.
1.18 Calculate the absolute and Celsius temperatures of a substance at 86 ◦ F.
(545.67R, 303.15K, 30 ◦ C)
1.19 Calculate the Celsius, Fahrenheit, and absolute (Rankine) temperatures of a
substance at 338K.
1.20 Calculate the Celsius, Fahrenheit, and absolute (Kelvin) temperatures of a
substance at 338R. (187.78K, −85.37 ◦ C, −121.67 ◦ F)
1.21 Calculate the absolute and Celsius temperatures of a substance at 338 ◦ F.
1.22 Calculate the absolute and Fahrenheit temperatures of a substance at 338 ◦ C.
(611.15K, 1100.07R, 640.4 ◦ F)
1.23 Normal body Celsius temperature is 37 ◦ C. What is the equivalent Fahrenheit
temperature (measured in ◦ F)?
1.24 At what body temperature is its Fahrenheit temperature value (measured in
◦
F) equal to the Celsius value (measured in ◦ C)? (−40 ◦ F)
1.25 Iron melts at 2795 ◦ F. How many times hotter is melting iron than melting ice?
1.26 A sequence of measurements of absolute pressure in a gas thermometer pro-
duces 16.2, 12.0, 8.70, 5.40 psi when the corresponding ice point absolute pressures
are 10, 8, 6, 4 psi. Use some plotting software with the best straight line fit, to obtain
the ideal gas temperature, and corresponding Fahrenheit temperature given by this
set. (322.04K, 120 ◦ F)
1.27 At what Fahrenheit temperature is a body twice as hot as melting ice.
1.28 Copper melts at 1083 ◦ C. How many times hotter is melting iron than melting
copper? (1.33)
Section 1.3.11
1.29 Consider the coefficient of thermal expansion of a thermometric liquid to be
a constant, α0 , when measured using a constant volume gas thermometer. Show by
integrating Eq. (1.51) using this constant value, and the initial condition
v = vi = Vi /M = (Vb + Azi )/M when T = Ti
that
Vb
z − zi = zi + [eα0 (T −Ti ) − 1]
A
Here Vb is the volume of the bulb, A is the cross-sectional area of the tube, and M is
the mass of the thermometric substance.
1.4 Exercises 45
1.30 Use the result of the previous problem in Eq. (1.45), to obtain the dimensionless
temperature y = (Tzg − Tig )/(Tsg − Tig ) as a function of the dimensionless ideal gas
temperature x = (T − Ti )/(Ts − Ti ). [Note that Example 1.2, is similar to this, and
that from Eq. (1.49) Ts − Ti = Tsg − Tig .].
1.31 Use the result of the previous problem
y = (ecx − 1)/(ec − 1)
with c = α0 (Ts − Ti ) to plot the calibration curve y − x as a function of x in the

range 0 ≤ x ≤ 1.2, with ethyl alcohol (α0 = 610 · 10−6 1/R) as the thermometric
substance. Use some plotting software to solve this problem. What is the maximum
Fahrenheit temperature difference between the readings of the two thermometers?
At what Fahrenheit temperature does it occur? (2.5 ◦ F, 122 ◦ F)
1.32 Differentiate the expression for the difference between thermometer readings
y − x using the function you plotted in the previous problem. Determine the max-
imum difference, Tzf − Tf , and the location of the maximum, Tf . Evaluate these
expressions using mercury and ethyl alcohol as thermometric fluids. For ethyl alco-
hol, the results should agree with what you found previously.
Chapter 2
Equilibrium
2.1 Introduction
In the previous chapter, we noted the role of equilibrium in the measurement of

pressure and temperature. Indeed, the concept of equilibrium is a central element in
thermodynamics where, as in mechanics, it provides a starting point for more general
considerations.
The notion of static equilibrium in mechanics is associated with the absence of
motion (relative to an inertial reference frame). The importance of static equilibrium
is that many practical systems are constrained, by external forces, to be at rest.
Moreover, experience shows that any system, which is started into motion in some
way but subsequently unforced, will ultimately come to rest. It is a fundamental
principle of mechanics that in a state of rest, or static equilibrium state, all properties
of a mechanical system are uniquely specified by the equations of statics along with
equations of state. The equations of statics are universal, they are the same for all types
of systems, however, different equations of state describe different material behavior
under load, and require different types of analysis. You have already learned about
the equilibrium of rigid bodies and linear elastic solids in your courses on statics and
strength of materials.
In thermodynamics, we deal with continuous collections of matter whose com-
plete description requires the specification of temperature as well as other properties.
We call such bodies, or collections of bodies, thermodynamic systems. The simplest
type of thermodynamic system consists of a single chemical constituent in a physi-
cally identifiable phase. We call this a pure substance.
The notion of thermodynamic equilibrium in thermodynamics is a generalization
of static equilibrium from mechanics. For a pure substance, it combines thermal
equilibrium, which is defined by an absence of heating, with static equilibrium, which
is defined by an absence of motion. These conditions apply to every element of the
substance. The importance of thermodynamic equilibrium is the same as I described
for static equilibrium, namely that many practical thermodynamic systems are found
in both constrained and unconstrained equilibrium states. Generalizing from the

48 2 Equilibrium
static case we note that in a thermodynamic equilibrium state, all properties of a

pure substance are uniquely specified by the conditions of no heating and no motion,
along with equations of state. Several important observations need to be made here:
• The condition of no heating is sufficient for the specification of thermal equilib-
rium, however, it is inadequate when a process occurs. This is different from the
situation with motion, where the condition of no motion produces the equations of
static equilibrium as a special case of the momentum evolution equations. How-
ever, the first law of thermodynamics, which we will study and you will learn to
use, establishes the energy evolution equation on an equal footing with momentum
evolution.
• The number of intensive properties that define the equilibrium of a pure substance
is equal to the number of equations that specify the equilibrium. This follows from
the fact that, at equilibrium, all properties of an arbitrary element are uniquely
specified; in thermodynamics this is called the state principle.
Although the solution of the equilibrium problem can be complex for solid bodies,
for fluid bodies it is relatively simple. This is fortunate, because fluids, particularly
gases, display the most pronounced thermodynamic behavior, and consequently our
study is largely focused on this class of substance.
2.2 Thermostatics of Pure Fluids
Thermal equilibrium, specified by the condition that the heating of an arbitrary fluid
element is zero,
Q̇ = 0 (2.1)
requires that the temperature of each element of the fluid is equal to every other (if
this were not so, then one element would heat another). Thus Eq. (2.1) implies that
T (x, t) = T (2.2)
where T is uniform in space and time. The equation of static equilibrium, as a special
case of no acceleration in Eq. (1.26), is
F+G=0
This can be applied to an arbitrary volume element of a continuous fluid using

Eq. (1.36) for the external force, F, Eq. (1.37) for the traction, and Eqs. (1.25) and
(1.29) for G [with Eq. (1.22) for the mass]. The result is

− p(x)n(x) d A − kg ρ(x) d V = 0
A V
2.2 Thermostatics of Pure Fluids 49
Fig. 2.1 Free body diagram of a fluid element. The equilibrium equation for the z direction makes
use of Taylor’s theorem
Applying this to a differential rectangular element (see Fig. 2.1) shows that the
pressure is a function of vertical location only, and in that direction it satisfies the
differential equation
dp g
p ≡ = −ρg = − (2.3)
dz v
Since there are two equilibrium equations in this case [Eqs. (2.1) and (2.3)], two
variables determine all the rest. If we take these as v and T , the formulation is
completed by the equation of state, in the form
p(x) = p[v(x), T (x)] (2.4)
I have written the argument x here to emphasize the fact that this equation is a local
relation among local (intensive) properties. The global solution of this, thermostatic
equilibrium, problem [Eqs. (2.1), (2.3), (2.4)] depends on the particular function
represented in Eq. (2.4), however, if we restrict our attention to a small vertical
distance, z, in which ρ does not vary significantly from ρ0 , the place where the
pressure is p0 , the integral of Eq. (2.3) is approximately
p ∼ p0 − ρ0 gz (2.5)
This is the well known hydrostatic pressure distribution, which is the basis of the
manometric pressure measurement scheme that we discussed in chapter 1. Moreover
if the vertical extent of the substance is restricted so that ρ0 gz/ p0 << 1 then neces-
sarily the fractional pressure variation, ( p − p0 )/ p0 , will be small. In this case we
can neglect even the second term on the right hand side of Eq. (2.5) and write, to a
high degree of accuracy,
p(x) ∼ p (2.6)
50 2 Equilibrium
This result means that the pressure is, for practical purposes, uniform throughout the
system. Based on Eqs. (2.2) and (2.6), we introduce the following definition.
Definition 2.1 (Macroscopic Equilibrium) A pure thermodynamic system exists
in a macroscopic equilibrium state when all its intensive properties are spatially
uniform, and are not changing in time.
Although we have just seen that macroscopic equilibrium so defined is an idealization
(at least in the vicinity of the earth due to the effects of gravity), it leads us to the
concept of a quasi-equilibrium state in which the variation of each intensive property
is so small it can be neglected; for example,
( pmax − pmin )/ pmin << 1
Many actual thermodynamic systems satisfy this sort of condition; the fractional
pressure difference between the floor and ceiling of an 8-foot high room full of air
at 70 ◦ F, is 0.000283 (0.03%).
In a macroscopic equilibrium (or quasi-equilibrium) state, each of the properties
appearing in Eq. (2.4) is a single number, and we write the equation without arguments
p = p(v, T ) (2.7)
The specific form of this equation completely describes the fluid behavior at equi-
librium. When we can solve this explicitly for v, we obtain the equation of state in
the form
v = v( p, T ) (2.8)
where p and T are the independent variables. In the third permutation, the fluid
pressure and specific volume determine the equilibrium temperature through the
equation of state in the form
T = T ( p, v) (2.9)
Equations (2.4), (2.8) and (2.9) are just different ways of writing the same equation of
state, by making explicit each of the properties involved. For example, the expression
p = C T /v, which is in the form of Eq. (2.7), can also be written in the form of
Eq. (2.8), v = C T / p, and Eq. (2.9), T = pv/C. Later when we introduce new
properties and their corresponding equations of state, we will refer to Eq. (2.7) as
the mechanical equation of state in order to distinguish it from the others.
Although in the preceding discussion we only considered pure fluids, there are
some other materials, and conditions, for which thermodynamic equilibrium can also
be specified by two numbers representing two of the three properties, p, v, and T . We
call them simple compressible substances, and because of their utility, we will focus
our study on them. One common type of simple compressible substance consists
of a combination of chemical elements which do not chemically interact; air is an
example that we will consider often. Another type is a solid under special loading; a
bar under a uniaxial load and a cube under an isotropic load are two examples.
2.3 The State Surface 51
2.3 The State Surface
Equation (2.7) represents a surface in the three-dimensional space of properties

p, v, T . This leads us to a simple geometric representation of a macroscopic equi-
librium state (a point on the surface) and a quasi-equilibrium state (a small cluster
of points on the surface) of a simple compressible substance. An example of such
a surface is shown in Fig. 2.2. The geometry of surfaces is somewhat more compli-
cated than the geometry of curves. It is also more difficult to visualize, and harder to
sketch. For this reason, we seek alternative means of representing the state surface.
If we set T to a constant value T1 , Eq. (2.7) becomes the equation of a curve, which
is the intersection of the state surface with the plane T = T1 . Selecting a different
value for T will produce a different curve. All these curves, which for obvious
reasons are called constant temperature or isothermal curves, can be plotted in the
same plane, say the coordinate plane T = 0 (we will also refer to this as the p, v
plane); the resulting picture is a contour plot of the surface with isothermal contours.
This gives us a way of visualizing the state surface in a plane. Figure 2.2 shows
this contour representation in addition to the surface itself. The constant temperature
curves are drawn on the surface as well as on the contour map. You may be familiar
with contour maps from cartography, where they are used to plot surface elevations,
and meteorology, where they are used to plot curves of constant atmospheric pressure.
We could also set p to a constant value in Eq. (2.6) and obtain a constant pressure
or isobaric curve. A set of isobaric curves plotted in the p = 0, or T, v plane is
another contour representation of the state surface. A third contour representation is
obtained by plotting a set of constant v curves in the v = 0, or p, T plane. These
are constant volume, or isochoric, curves. These two types of representation are
illustrated in Fig. 2.3 for the surface shown in Fig. 2.2. The isochors are drawn on
the surface as well as in the p, T plane. Anyone of these different contour plots can
be used as an aid in visualizing the state surface.
Fig. 2.2 An example of an equilibrium state surface and one of its contour representations
52 2 Equilibrium
Fig. 2.3 Contour maps for the surface shown in the previous figure
2.3.1 The Geometry of Curves
In this section we review some of the facts about curves that you need to know for this
course in thermodynamics, as well as other technical courses that you are studying.
A function, y = y(x), continuous and having a continuous derivative,
dy
= Dx y
dx
on an interval of the x axis, is a representation of a smooth plane curve. The physical
meaning of continuity here is that there are no jumps in either the curve or its slope.
It is often convenient to represent a plane curve, C, in terms of a parameter. This
is done by means of the equations
y = yC (t) x = xC (t) (2.10)
For every value of t we obtain from these a value for x and of y, and thus in a
specific parametric representation every point on the curve is associated with a value
(possibly more than one) of t. A parametrization is not unique; a single curve can
be represented by many different parametric relations. Moreover if we can solve the
second equation to obtain t = tC (x), and substitute into the first, we obtain the usual
cartesian form
y = yC [tC (x)] = y(x)
Example 2.1 Plot the curve that corresponds to the parametric representation
y = y0 + b sin 2πt/tc x = x0 − a cos 2πt/tc 0 ≤ t ≤ tc
and express it in cartesian form.

Solution. A plot of the curve is shown in the accompanying figure. In order to

express it in parametric form we must eliminate t from the equations. This is
done most easily in the present case by isolating the trigonometric function in
each equation, squaring, adding the squares, and using a well-known identity.
The result is
(y − y0 )2 /b2 + (x − x0 )2 /a 2 = 1
This is the equation of an ellipse centered at x0 , y0 with semi-axes a and b.

In the curve drawn several of the points are marked with an appropriate value
of, t/tc . However, note also that each point on this curve corresponds to many
values of t.
Differentiation of a parameterized curve is accomplished by using the chain rule.

For a curve y = y(x) described by Eqs. (2.10), so that
y = y[xC (t)] = yC (t)
this is
dyC dy d xC d xC
= = Dx y (2.11)
dt d x dt dt
When Eqs. (2.10) are given explicitly the derivative can be evaluated from this
−1
dy dyC d xC
=
dx dt dt
provided that d xC /dt = 0; however, when Dx y is given implicitly as a function of

x and y, Eq. (2.11) is a differential equation that relates yC (t) to xC (t). In the special
case Dx y = f (x), Eq. (2.11) can be integrated. Using points O and P on the curve,
corresponding to the values t O and t P , the result takes the form
tP
d xC
yC (t P ) = yC (t O ) + f [xC (t)] dt
tO dt
With a change in variable of integration (but this can be done only on a t interval for
which d xC /dt is continuous and different from zero) this is simply
54 2 Equilibrium
xP
y P = yO + f (xC ) d xC
xO
which, since P is an arbitrary point on the curve, can be regarded as the cartesian
form of the curve passing through O and having the specified derivative
x
y(x) = y O + f (xC ) d xC (2.12)
xO
This should emphasize the fact that a curve can be specified in this way. Indeed in
thermodynamics, this is very often the way curves are given.
If we set f (xC ) equal to a constant, f (x O ), Eq. (2.12) becomes the equation of
the tangent line to the curve at the point O. Moreover for this simple f , the integral
can be evaluated and the curve is given explicitly by
yT (x) = y O + Dx y(x O )(x − x O )
where the variable yT is the value of the ordinate on this tangent line, and Dx y(x O )
means the derivative evaluated at x O .
If Dx y(x O ) is approximated by (y P − y O )/(x P − x O ) where x P , y P is a specific
point on the curve, Eq. (2.12) becomes the equation of the chord or secant line
between the two points; in integrated form

y P − yO
yS (x) = y O + (x − x O ) (2.13)
xP − xO
where yS is the value on the secant line connecting the points. The relations among
a curve, its tangent, and its secant (chord) are shown graphically in Fig. 2.4. You can
see in the figure that if the change in slope of the curve between P and O, is not
too large, the tangent, and the chord will both be good approximations to the curve
in the interval. We will use the chord in this way to obtain approximate values on a
curve between two known values, a procedure called linear interpolation.
Example 2.2 Given a curve in the following tabular form
x 0 0.2 0.4 0.6 0.8 1.0

y 0 0.04 0.16 0.36 0.64 1.0
determine the value of y at x = 0.7 by linear interpolation.

Solution. In applying Eq. (2.13) to this example we choose the points O and
P to be 0.6 and 0.8, respectively, because these are the closest values to the
required 0.7 and bracket it. Then we obtain
yS (0.7) = 0.36 + [(0.64 − 0.36)/(0.8 − 0.6)](0.7 − 0.6) = 0.5

Fig. 2.4 The geometric

relations between a plane
curve, its tangent line at a
point, and a chord line
through the point
The actual value is y(0.7) = 0.49 (the table was created using the function
y = x 2 ) so the error made by replacing this value by yS (0.7) in this case is
about 2%.
A curve in space is denoted by the parametric form
x = xC (t) y = yC (t) z = z C (t) (2.14)
Any pair of these, for example,
x = xC (t) y = yC (t)
is a plane curve C̄, which is the projection, or shadow, of C in the appropriate plane
(here the x, y plane). Eliminating t between these by inverting one of them, t =
tC (xC ), and substituting into the other, y = yC [tC (xC )] = yC̄ (xC ), gives the equation
of the projection in cartesian form.
In the event that the parameter is time, the derivatives of the functions in Eq. (2.14)
are the components of velocity (see Sect. 1.3.4).
2.3.2 The Geometry of Surfaces
As mentioned earlier it can be difficult to visualize geometric relations on a surface,

because three dimensions are involved. Moreover because of the added dimension,
the geometry of surfaces is somewhat more complex than that of curves. However,
56 2 Equilibrium
in order to try to make this review as clear as possible for you, I have arranged the
following discussion so that it is similar to the previous one on curves. If you are
confused at someplace, try looking back at the previous section about curves.
A continuous function z = z(x, y) in a region of the x, y plane with continuous
first partial derivatives
∂z ∂z
≡ ∂x z ≡ ∂y z (2.15)
∂x ∂y
and second partial derivatives
∂2z ∂2z ∂2z ∂2 z

(2.16)
∂x 2 ∂x∂ y ∂ y∂x ∂ y2
is a representation of a smooth surface1

A theorem of Calculus, which is especially important in thermodynamics states
that if both derivatives appearing in Eq. (2.15), and one of the mixed partial derivatives
appearing in Eq. (2.16) are continuous in some region of the x, y plane, then
∂2z ∂2z
= (2.17)
∂x∂ y ∂ y∂x
everywhere in that region. Note, however, that this equation has no counterpart in
the geometry of curves.
Example 2.3 Calculate the first and second partial derivatives for the function
RT
p=
v
in the region v ≥ 0, T ≥ 0. The quantity R is a dimensional constant; the
surface of Fig. 2.2 is a plot of this function.
Solution. Doing the indicated differentiations gives for the first partial
derivatives
∂p ∂p
= R/v = −RT /v 2
∂T ∂v
Differentiating these expressions gives the second partial derivatives as
∂2 p ∂2 p
=0 = −R/v 2
∂T 2 ∂v∂T
1 Each partial derivative is done holding the other variable constant, but we will not represent this
explicitly as we did (see Sect. 1.3.11) as long as there can be no confusion.
and
∂2 p ∂2 p
= −R/v 2 = 2RT /v 3
∂T ∂v ∂v 2
Except for v = 0, all quantities are continuous and the two mixed partial deriva-
tives are equal as Eq. (2.17) requires. Note that at v = 0, T = 0 the surface
has a corner (see Fig. 2.2); there is no unique tangent plane there.
For a space curve, defined by Eqs. (2.14), that lies in a surface z = z(x, y), we
have
z = z[xC (t), yC (t)] = z C (t)
so applying the chain rule gives

dz C ∂z d xC ∂z dyC d xC dyC
= + = ∂x z + ∂y z (2.18)
dt ∂x dt ∂ y dt dt dt
Like its plane counterpart, Eq. (2.10), when the two partial derivatives are known
implicitly as functions of x, y and z, this is a differential equation relating the three
parametric functions. On writing
∂x z = A(x, y) ∂ y z = B(x, y) (2.19)
we can integrate Eq. (2.18) from O to P along C

tP
d xC dyC
z C (t P ) = z O + A[xC (t), yC (t)] + B[xC (t), yC (t)] dt (2.20)
tO dt dt
With a change in the variables of integration (see Sect. 2.3.1) the projection C̄ of C
can be expressed in cartesian form and written as
xP yP
z C (x P , y P ) = z O + A[xC , yC̄ (xC )] d xC + B[xC̄ (yC ), yC ] dyC (2.21)
xO yO
The value of the integral is the difference in the two surface heights, z P − z O , inde-
pendent of which curve C is used to effect the integration; any curve lying in the
surface will do, as you can see in Fig. 2.5. The information that the curve lies in the
surface is contained in Eq. (2.21) through Eqs. (2.17) and (2.19) in the form
∂ y A = ∂x B (2.22)
Consequently the equation of the surface can be expressed, by using the curve whose
projection is C̄ x and C̄ y as shown in Fig. 2.5, and dropping the subscript P, since it
refers to an arbitrary point in the surface
x y
z(x, y) = z O + A(xC , y) d xC + B(x O , yC ) dyC (2.23)
xO yO
58 2 Equilibrium
Fig. 2.5 An illustration showing the space curves, corresponding to the projections C̄ and C̄ x + C̄ y ,
that lie in the continuous surface z(x, y)
In this way the equation of the surface passing through O is expressed in terms of its
first partial derivatives as can be seen from Eq. (2.19). If we set the partial derivatives
equal to constants A(x O , y O ), and B(x O , y O ), we obtain the equation of the tangent
plane to the surface at O. On evaluating the integrals in this simple case we obtain
z T − z O = ∂x z(x O , y O )(x − x O ) + ∂ y z(x O , y O )(y − y O )
where z T is the value in this plane.

Alternatively if the partial derivatives appearing in Eq. (2.23) are approximated
by (z P − z O )/(x P − x O ) and (z Q − z O )/(y Q − y O ), where the subscripts denote
three points on the surface, then integration produces the equation of the chord plane
passing through these points

zP − zO zQ − zO
zS − zO = (x − x O ) + (y − y O ) (2.24)
xP − xO yQ − yO
where z S is the value in the chord plane determined by the three points. All these topics
are generalizations from the previous section on curves. However, in the present
instance, a situation can arise which has no counterpart in the geometry of curves.
Since this occurs in thermodynamics, we include it here.
Suppose we are given a plane curve, C̄, described as in Eq. (2.10), and a differential
equation
dx dy
ż = A(xC̄ , yC̄ ) C̄ + B(xC̄ , yC̄ ) C̄ (2.25)
dt dt
Fig. 2.6 The sketch on the left shows a surface element for which the mixed partial derivatives at
the origin are equal. The element on the right, for which they are unequal, is discontinuous, and the
value of z(x P , y P ) depends on the path chosen
where ∂ y A = ∂x B in some region of the (x, y) plane. Then, although the integral
is still given by Eq. (2.20) or (2.21), its value depends on the plane curve C̄ along
which the integration was performed z = z(x, y, C̄) (see the sketch on the right
of Fig. 2.6), and there is no continuous surface, z = z(x, y), which is comprised
of all the space curves (x(t), y(t)z(t, C̄)). In such as case we call the integral an
interaction. Figure 2.6 illustrates the connection between continuity of the surface
and the equality of the mixed partial derivatives. Note that for a continuous surface
the angles θx and θ y must be equal, and these are the change of y slope in the x
direction, ∂x ∂ y z, and the change of x slope in the y-direction, ∂ y ∂x z, respectively.
A final topic (we did not discuss this in the section on curves) relates to the
conditions under which we can explicitly solve the state equation for another of its
variables. Consider a point for which ∂x z is not equal to zero. Then Eq. (2.18) can
be solved for d xC /dt, and we find
d xC 1 dz C ∂ y z dyC
= −
dt ∂x z dt ∂x z dt
This is the equation for the derivative of a space curve that lies in the surface described
by the inverse function x = x(z, y) and consequently, it can also be written as
d xC dz C dyC
= ∂z x + ∂y x
dt dt dt
Since these two expressions for d xC /dt must be the same for all possible values of
dz C /dt and dyC /dt, we must have
1 ∂y z
∂z x = ∂y x = − (2.26)
∂x z ∂x z
These purely mathematical relations among partial derivatives have important physi-
cal meaning when they occur in the context of thermodynamics. If the partial deriva-
tives in Eq. (2.26) are known in terms of z and y, the surface can be constructed from
60 2 Equilibrium
them using Eq. (2.23). This discussion shows that a necessary condition to be able to
solve the equation z = z(x, y), explicitly for x = x(z, y) in the vicinity of the point
O, [refer to Eq. (2.8)], is that
∂x z(x O , y O ) = 0
This condition is satisfied for almost all the thermodynamic state points of interest
to us in this course, consequently we will invert our state equations without further
discussion. Where it is not satisfied, we will note those singular points.
2.3.3 Thermostatic and Thermodynamic Problems

Using the equation of state along with the definition of specific volume, Eq. (1.23),
we can determine two of the five properties, p, v, T, V, M, from a knowledge of
the other three. However these three must be given either as a combination of any
two intensive and one extensive property, or as a combination of one intensive (not
specific volume) and two extensive properties. This can all be summarized in the
following tables.
State Properties
Equations
M p
p = p(v, T )
V T
V = Mv
v
We can solve many practical thermostatic problems using this system.
Example 2.4 A mass M of a simple compressible substance exists at a gauge

pressure of 25.3 psi and a temperature of 76 ◦ F. The atmospheric pressure is
14.7 psi, what volume does the substance occupy?
Solution. The information given in all problems of this type can be organized
similarly to the way I have done here. Note that a temperature given in ◦ F is a
Fahrenheit temperature (and likewise ◦ C denotes a Celsius temperature)
State 1
M1 = M Tf 1 = 76◦ F
p1 = 40 psi
Of the five properties, three are given, either symbolically as mass is given
here, or numerically. Of course, numerical values must always be accompanied
by an appropriate unit. The remaining two properties can be determined using
the equation of state and definition of specific volume. In the present problem
the equation of state in the form Eq. (2.8) allows us to calculate the specific
volume as
v1 = v( 40 psi, 536 R)
and subsequently Eq. (1.23)

V1 = Mv1
produces the volume occupied by the substance.
A more interesting class of problems involves two equilibrium states described

by ten properties. Of the ten, six can be specified, and four determined. The specified
properties are distributed three in each state, however, the six specified properties
may not all be stated explicitly; they can be implied by the problem description. The
following example illustrates these points for this two-state class of problems. These
are thermodynamic problems in the sense that the two equilibrium states are related
by a thermodynamic process. Processes are driven by interactions such as heating a
system or doing work on it. They also occur when a constraint is removed, and the
system moves to another equilibrium state.
Example 2.5 An 18 ft3 rigid vessel contains a simple compressible substance

at 70 ◦ F and 20 psi. If the substance is heated until the final pressure is twice
the initial, what is the final temperature? Sketch this change of state on an
appropriate contour plane of the state surface.
Solution. All two-state problems should be solved as I indicate in this problem.
First, organize the given information by state as shown below. You must include
a reason for everything you write, except when you write numerical values; you
can find all the information you need, along with the reasons, by reading the
problem statement. Place a sketch of the system, as it appears during the change
of state, between the state information. This sketch is the analog of a free body
diagram in statics; it should include the interactions that are occurring, and
be detailed enough to provide a sense of what is physically going on in the
problem.
State 1 State 2
V1 = 18 ft3 Tf 1 = 70◦ F p2 = 2p1 M2 = M1
(stated) (fixed mass)
p1 = 20 psi V2 = V1
(rigid vessel)
You must not continue beyond this point until you have written six equa-
tions, as I have here. If you have not written six equations, you do not have
enough information to solve the problem completely; go back and read the
problem statement again—more carefully and more critically. Note that if the
pressure is not specified as gauge pressure, you may take it to be the absolute
pressure. The equations you have written should be distributed, as they are
62 2 Equilibrium
above, three in each state. By using the equation of state in the appropriate
form, and the definition of specific volume, for each state, we can determine
all thermodynamic information. In the present problem we are asked for the
final temperature, which can be expressed by Eq. (2.9) as
T2 = T ( p2 , v2 )
The extensive equations of state 2 imply that the specific volume is unchanged
v1 = V1 /M1 = V2 /M2 = v2
so the change of state can be easily sketched on a p, v plane. The final

temperature can be written in terms of the state 1 properties
T2 = T (2 p1 , v1 )
and, although v1 is not known, it can be calculated using Eq. (2.8)
v1 = v(20 psi, 530R)
which means that the final temperature is calculable
T2 = T (40 psi, v1 ) = T [40 psi, v(20 psi, 530R)]
Note that we did not use the information given about V1 ; it could just as well
not have been given. Some problems can be incompletely specified in this
sense, and we can still solve for all the intensive properties.
The following example illustrates the need to recall, and use, ideas that you learned
previously in statics and elsewhere.
Example 2.6 A mass M of a simple compressible substance at pressure p and

temperature T is contained in a cylinder fitted with a piston of area A. The
substance is heated until the piston is twice as high as it was initially. What is
the initial height of the piston, z 1 , what is the mass of the piston, m p , and what
is the final temperature of the substance? The atmospheric pressure is pa , no

substance leaks past the piston, and the static friction is negligible compared
with the other forces acting on the piston.
Solution. Organize the given information as before
State 1 State 2
M1 = M T1 = T p2 = p1 M2 = M1
(piston (no leakage)
p1 = p equilibrium) V2 = 2V1
(geometry)
The equation for V2 was obtained by using the geometrical formula for
the volume of a cylinder V = z A, along with the information given about the
height ratio. The equation for p2 was obtained by observing that at both states 1
and 2, the piston is in static equilibrium. Thus, according to the accompanying
free body diagram, which neglects static friction (see Example 1.4)
p A − pa A − m p g = 0
so that
p1 = p = pa + m p g/A = p2
This equation can also be used to determine the piston mass in terms of the
other given quantities
m p = ( p − pa )A/g
Since the final specific volume is twice the initial
v2 = V2 /M2 = 2V1 /M1 = 2v1
and since the initial specific volume can be calculated from the given data
using Eq. (2.8)
v1 = v( p, T )
64 2 Equilibrium
we can calculate the final temperature using the form, Eq. (2.9)
T2 = T ( p, 2v1 ) = T [ p, 2v( p, T )]
Moreover we can use Eq. (1.23) along with geometry V1 = v1 M = h 1 A to

obtain a calculable expression for the initial height
z 1 = v1 M/A
The state points are best indicated on a p, v plane as shown above.
We can also solve three state problems, they relate 15 properties of which 9 are given
and 6 need to be determined, as well as problems in which the specified properties do
not explicitly define either state, and other variations. I will give examples of some
of these later, but first I want to introduce some equations of state, so that we can
solve real problems involving real substances.
2.4 The Mechanical Equation of State

The most direct determination of the equation of state of any fluid is obtained by
simply measuring triples of values of p, v, T and compiling the results in an organized
form. However we can learn more by considering the geometry of the surface. As
we saw in Sect. 2.3.2, when a space curve
p = pC (t) v = vC (t) T = TC (t)
lies in the state surface, p = p(v, T ), its derivative is

dpC ∂ p dvC ∂ p dTC
= +
dt ∂v dt ∂T dt
and in the neighborhood of a point for which ∂ p/∂v is continuous and not equal
to zero, we can solve this for dvC /dt. The result can be expressed in terms of the
compressibility, Eq. (1.53), β = −∂ p v/v, by using Eq. (2.26) in the form, ∂ p v =
1/∂v p, as
dvC ∂ p dTC dpC
= vC β − vC β
dt ∂T dt dt
Since this is the differential equation of a space curve that lies in the surface, v( p, T ),
the coefficient of the term, dTC /dt, must be ∂v/∂T . Then using the definition of the
coefficient of thermal expansion, Eq. (1.51), α = ∂T v/v, we write this equation in
the form
dvC dTC dpC
= vC α − vC β
dt dt dt
or on division by vC
d ln vC dTC dpC
=α −β (2.27)
dt dt dt
2.4 The Mechanical Equation of State 65
and also note, as a special case of Eq. (2.26), the important relation
∂p α
= ∂T p = (2.28)
∂T β
The coefficients in Eq. (2.27) satisfy the condition of continuity, Eq. (2.22) which
in this case is
∂ p α = −∂T β (2.29)
and if they are known as functions of p and T , then Eq. (2.27) can be integrated [as
in Eq. (2.23)] to give the equation of state

v T p
ln = α( p O , TC ) dTC − β( pC , T ) dpC (2.30)
vO TO pO
One common way that Eq. (2.29) is satisfied is when α is a function of temperature
only and β is a function of pressure only.
2.4.1 Liquids
The preceding analysis is correct for all states of matter, but it is most useful with
liquids, for which simple approximations of the equation of state can be obtained
by making use of their relative incompressibility. We illustrate this in the following
sections.
Constant Volume Liquids
If α and β are both zero, Eq. (2.30) produces the equation of state
v = vO
This state surface is a plane parallel to the p, T plane; arbitrary values of pressure
and temperature correspond to the same specific volume. Since the fluid does not
change its volume under any circumstance, we say that such a material has constant
volume. A constant volume substance is different from a rigid body. The latter cannot
deform in any way, whereas the former can change shape provided that its volume
remains fixed; a constant volume substance by definition is incompressible. This
simple equation of state is a very useful thermodynamic approximation for liquids
(and solids).
Constant Property Liquids
A more realistic equation of state for a liquid is obtained when α and β are nonzero
constants. Then Eq. (2.30) produces
66 2 Equilibrium
v = v O e[α O (T −TO )−β O ( p− p O )]
Measurements on real fluids show that both α O TO and β O p O are extremely small for
ambient values of pressure and temperature. Consequently, for small fractional pres-
sure and temperature changes (up to about 100), the linearized expression, obtained
by expanding the exponentials,
v − vO
∼ α O (T − TO ) − β O ( p − p O ) (2.31)
vO
is highly accurate. This, equation of state of a linear elastic fluid (or solid), also is
represented by a plane surface (the tangent plane to the actual surface at O). However,
in this case, the surface is inclined to the coordinate axes. Of course if both α O TO
and β O p O are small enough to be ignorable, Eq. (2.31) is simply the equation of
state of a constant volume fluid. Another possibility which reflects the real situation
for many fluids is,
β O ( p − p O ) << α O (T − TO ) << 1
so that Eq. (2.31) can be further approximated as
v − vO
∼ α O (T − TO ) (2.32)
vO
Since here we have effectively β ∼ 0, the volume of such a fluid can change only
due to temperature changes, not by mechanically compressing it. It is a mechanically
incompressible fluid.
Example 2.7 Glycerin exists at atmospheric pressure (101 kPa) and a tem-
perature of 18 ◦ C. If it is stirred, in a vat open to the atmosphere, until its
temperature is 46 ◦ C, what is the change in its specific volume?
Solution. The given information is organized in our form as follows:
State 1 State 2
Tc1 = 18◦ C p2 = p1 M2 = M1
(open vat) (fixed mass)
p1 = 101 kPa Tc2 = 46◦ C
Here, only five quantities are given so that we cannot determine all state
properties, although we can determine all the intensive properties. Here how-
ever we are asked only for the change in v, and therefore instead of numerically
applying Eq. (2.31) twice, once to each state, we first subtract the equation
applied to state 1 from its form when applied to state 2. This gives
v2 − v1 = v O α O (T2 − T1 ) − v O β O ( p2 − p1 )
The advantage in doing this is that the resulting equation does not contain either
TO or p O . Using this with the data provided, along with information from the
table of properties, gives [recall from Eq. (1.49) that T2 − T1 = Tc2 − Tc1 ]
v2 − v1 = (0.784 · 10−3 m3 /kg)(0.504 · 10−3 1/K)[(46 − 18) ◦ C · 1K/ ◦ C]
Accordingly, the change in specific volume is 11.1·10−6 m3 /kg.
Linear Elastic Solids

As we mentioned previously, solids are not simple compressible substances. In gen-
eral there are six components of stress that depend on six components of strain in
addition to temperature. Moreover, static equilibrium conditions do not result in spa-
tially uniform stresses. However for bars of uniform cross section under uniaxial
load, there is only one nonzero, uniform, component of stress which only depends
on the extensional strain and temperature. Thus in this special case, the solid behaves
like a pure fluid in macroscopic equilibrium, and is an example of a simple compress-
ible substance, namely a substance and loading for which two independent variables
specify a global state.
An equation of state that relates the stress to extensional strain, can be derived by
reasoning as we did with Eq. (2.31),
x − xO σx
x = ∼ α (T − TO ) + β σx = α (T − TO ) + (2.33)
xO E
This is Hooke’s law that you studied in Strength of Materials. In this expression,
Young’s modulus and the coefficient of linear expansion are defined slightly differ-
ently than α and β, but they all can be related by using the other equations of state,
which involve the transverse strains
1 − 2ν
α O = 3α βO =
E
Here ν denotes Poisson’s ratio, which is a measure of the contraction of the cross-
sectional area as the material extends. In this course, our interest is principally in the
thermal behavior of liquids and solids, and for these purposes the equation of state of
an incompressible substance, v = v O , is often sufficient. Although we will not use
Eq. (2.33), I have written it here so you can see the relationships among the things
you are learning here.
Vaporizable Liquids
A characteristic of liquids is that they obey an equation of state like Eq. (2.31) only
over a limited range of specific volume. Typically as the temperature of the liquid is
increased at constant pressure, a limit is reached at which the liquid begins to boil.
The amount of liquid decreases and a corresponding amount of vapor appears. We
call this phenomenon a phase change. At a given temperature T , the limiting specific
volume (which depends on the temperature) is denoted by v f (T ). This is shown
68 2 Equilibrium
Fig. 2.7 Contour maps of the equation of state of a vaporizable liquid showing the saturated liquid
curve. The path of integration used for v( p, T ) in this section is shown in gray in the p, T plane
in Fig. 2.7; the limiting curve is called the saturated liquid curve. The isobars in
the figure are almost vertical (their slope is 1/v O α O ), therefore I have drawn them
with an enlarged horizontal scale so you can visualize them more easily. As you
can also see in the figure, each point on the saturated liquid curve corresponds to a
unique pressure. Therefore there is another function, the saturation pressure ps (T ),
which relates the pressure at which the liquid boils to the boiling temperature. The
corresponding, saturation curve, is shown on the p, T plane sketched in Fig. 2.7.
This sketch distorts the slope of the isochors (curves of constant v), α O /β O [recall
Eq. (2.28)], which are almost vertical compared with the saturation curve, p = ps (T ).
By adapting the path of integration used in Eq. (2.30) to the one shown in Fig. 2.7,
the liquid equation of state is expressed in terms of the functions v f (T ) and ps (T )
that define the saturated liquid curve
p
v( p, T ) = v f (T ) exp − β( pC , T ) dpC p ≥ ps (T )
ps (T )
in which v f (T ) is
T ps (T )
v f (T ) = v O exp α[ ps (TC ), TC ] dTC + β[ pC , Ts ( pC )] dpC
TO pO
and T = Ts ( p) is the inverse of p = ps (T ), an alternative expression for the satura-

tion curve.
Now if we recall, from our previous discussion of liquids, that the exponential
involving pressure in the equation is close to one, we can drop it with little loss of
accuracy and obtain
v( p, T ) ∼ v f (T ) p ≥ ps (T ) (2.34)
This is the equation that is normally used as the equation of state of a vaporizable
liquid. It is used directly when p and T are given as the independent variables, and
is inverted to find T when v and p are specified
T ( p, v) ∼ T f (v) p ≥ ps [T f (v)] = p f (v)
As we have already seen, such a liquid is mechanically incompressible, however its

volume can change with temperature.
Data on v f , and ps , have been measured and tabulated at various temperatures
for many substances; the tables provided to you contain this information for water,
a substance important in power producing applications, and R-12, a substance used
in some refrigeration applications. Sometimes the saturation pressure, ps , is known,
and in terms of this variable the equation of state takes the form
v( p, ps ) ∼ v f [Ts ( ps )] = v̂ f ( ps ) p ≥ ps
You should use whichever form is more convenient for a specific problem.
Example 2.8 Four and a half pounds of water exists at a pressure of 18 psi
and a temperature of 73 ◦ F. What volume does the water occupy? The water is
heated at constant pressure until boiling begins. What is the water temperature
at this point? If the piston diameter is 6 in, what is its change in height?
Solution. The given information is organized in our form as follows:
State 1 State 2
M1 = Tf 1 = 73◦ F p2 = p1 M2 = M1
4.5 lbm (constant (closed)
p1 = 18 psi pressure)
x2 = 0
(boiling)
From the property table for water, state 1 is in the vaporizable liquid region,
however, the specific values of pressure and temperature used in this problem
do not appear in the table. Therefore we use the equation of the chord plane,
Eq. (2.24), to interpolate for v1

v P − vO vQ − vO
v1 − v O = (T1 − TO ) + ( p1 − p O )
T P − TO pQ − pO
There is no property variation with pressure, so this reduces to the chord line
of v f (T ) [see Eq. (2.34)]. Then using the v f values at the temperatures 60 ◦ F
and 80 ◦ F corresponding to P and O (these most closely bracket 73 ◦ F), we
obtain
70 2 Equilibrium

0.016033 + [(0.016072 − 0.016033)/(80 − 60)](73 − 60)
v1 =
0.016058 ft3 /lbm
According to the problem statement, the final state is on the saturated liquid
curve.
We show this, as an aid in visualizing the change of state, on the T, v
diagram to the right. In this sketch, we have shown the constant pressure curve
distinct from the saturated liquid curve. We have done this for visualization
purposes even though according to the approximation Eq. (2.34), these two
curves should appear as one in this plane. The final conditions are given by
the equation of state with the condition of boiling at the final pressure which
I have denoted by the equation x2 = 0, in other words knowing p2
x2 = 0 −→ v2 ∼ v̂ f ( p2 ) and T2 = Ts ( p2 )
However since this value of the saturation pressure does not appear in the
property tables, we must use Eq. (2.13) for the chord line to interpolate the
final temperature

Tf P − Tf O
Tf2 − Tf O = ( p2 − p O )
p P − pO
We can use either section of the saturation curve to do this, however, the
horizontal segment is easier to use because of the simple values of pressure
that appear

211.99 + [(227.96 − 211.99)/(20 − 14.7)](18 − 14.7)
Tf2 =
221.9 ◦ F

0.01672 + [(0.01683 − 0.01672)/(20 − 14.7)](18 − 14.7)
v2 =
0.01679 ft3 /lbm
The change in height is simply
z = (V2 − V1 )/A = (v2 − v1 )M1 /(π D 2 /4)
which can be calculated in terms of known quantities


(4.5 lbm)[(0.01679 − 0.016058) ft3 /lbm](1728 in3 /ft3 )/[π (3 in)2 ]
z=
0.20 in
The change in height is only written with two significant figures because of
the loss of three significant figures in the subtraction of the specific volumes.
2.4.2 Gases
Equations of state for gases cannot be obtained as we did for liquids, because gases
are much more compressible. However, in addition to measurement of properties
and tabulation, it is possible to gain some insight into their behavior through simple
particle models.
Ideal Gases
One of the simplest, yet useful, examples of an equation of state is provided by an
ideal gas, which is a model substance composed of noninteracting particles such as
we described in chapter 1. Accordingly its pressure and temperature are given by
Eqs. (1.41) and (1.51), respectively
2 m p c2 3 m p c2
pV = Np kB T =
3 2 2 2
Combining these gives the equation of state
pV = k B N p T
We do not use the equation in this form because it contains the kinetic theory quantity
N p ; rather, we introduce the dimensional quantity N representing a new dimension,
the amount of substance [N]. Then N can be measured in terms of any units that
describe quantity, such as pair, dozen, or gross (particles are a special case for which
N = N p particles). In terms of N = N p /N A (N A is a dimensional constant with
dimension [N −1 ]) the equation is
pV = kU N T (2.35)
where the universal gas constant, kU = k B N A , has the dimension [E/NT], and the
same value for all gases. The base unit for N in the SI, as defined by the CGPM, is
the mole (mol), given by the conversion relation
1 mol = 6.02214076 × 1023 elementary entities (2.36)

72 2 Equilibrium
The number in this equation2 is the fixed numerical value of the Avogadro constant,
N A when expressed in the unit mol−1 and is called the Avogadro number in honor
of Amadeo Avogadro, who was one of the early contributors to the atomic theory of
matter.
Equation (2.35) is in good agreement with the classical measurements made on
real gases over a wide range of pressure and temperature, by Boyle, Mariotte, Gay-
Lussac, Regnault, and Charles. Moreover, it expresses Avogadro’s atomic hypothesis
for gases; namely, that equal volumes of different gases, at the same pressure and
temperature, contain the same quantity of matter. When expressed in terms of the
molar volume, v̄ = V /N (note that this is similar to specific volume, but based on
quantity) it is especially simple
p v̄ = kU T.
Equation (2.35), called the ideal gas law because it is not exact for real gases, is
more useful to engineers when written on a mass basis. Then the molar mass
m̄ = M/N (2.37)
(defined like the molar volume above) is an important variable. The base unit of
molar mass is kg/mol; however, for historical reasons g/mol is most often used.
Other frequently used units are the kmol and lbmol, defined by
1 kg/kmol = 1 lbm/lbmol = 1 g/mol.
Based on the Atomic Theory of Matter, initiated by John Dalton, we can write the
molar mass of a substance whose relative molecular mass (molecular weight) is M
as
m̄ = M g/mol (2.38)
Multiplying and dividing Eq. (2.35) by mass M produces
pV = M RT (2.39)
where
R = kU N /M = kU /m̄ (2.40)
is a constant of dimension [E/MT], whose value depends on the specific gas con-
sidered, through its molar mass. The value of kU , obtained (see its equation on the
previous page) from the known values of the Boltzmann constant, k B in Sect. 1.3.9,
Kinetic Model for Temperature and the Avogadro number, Eq. (2.36), is the same
for all gases
2 An elementary entity may be an atom, a molecule, or any other particle or specified group of
particles. The amount of substance is a measure of the number of specified entities.
kU = 8.314463 kJ/kmol K = 1545.347 ft lbf/lbmol R
Values of R for specific gases can be obtained by using Eq. (2.40) and their molecular
weight. These numerical values are given in tables of thermodynamic properties for
a number of common ideal gases.
Equation (2.39) is the form of the ideal gas equation of state that is most used
by engineers. Using the definition of specific volume Eq. (1.23), V = v M, it can be
written in the intensive form
pv = RT (2.41)
The following example illustrates the use of the ideal gas equation of state. It is also
an example of a problem in which neither state is given explicitly.
Example 2.9 An 18 ft3 vessel contains 3 lbm of Methane. The gas is com-
pressed by a piston until the volume is half the initial volume, and the tem-
perature is twice the initial temperature. If the pressure has increased by
140 psi, and the atmospheric pressure is 14.7 psi, what did the pressure gauge
read prior to the compression? What is the final temperature?
Solution. The problem information is summarized as follows:
State 1 State 2
M1 = p1 = p2 − M2 = M1
3 lbm 140 psi (closed)
V1 = (stated) T2 = 2T1 V2 = V1 /2
18 ft3 (stated) (stated)
We cannot calculate T2 because we do not know either p2 or T1 , and likewise

for state 1 properties. However, if we were to guess a value for p1 , call it
p1G , then we could calculate everything else, including the specific volume in
state 2
v2G = v( p2G , T2G ) = v( p1G + 140 psi, 2T1G ) = v[ p1G + 140 psi, 2T ( p1G , v1 )]
Unfortunately this value will not be equal to the actual value
v2 = V2 /M2 = V1 /2M1 = v1 /2
there will be an error

ε = v2G − v2
74 2 Equilibrium
which can be calculated, and plotted, as a function of p1G . This function is

sketched in the accompanying figure. We can see from it that the correct initial
pressure is the one that corresponds to a zero error. Moreover we do not have
to make a large number of guesses, only one that gives a positive error and
another that gives a negative one. The actual pressure can then be approximated
by the chord line between the points as shown.
The situation is much simpler in the present problem because we can
approximate the behavior of Methane by Eq. (2.41), and the error equation
is
R2T1G 2 p1G v1 v1
ε= − v2 = −
p1G + 140 psi p1G + 140 psi 2
For ε = 0 the solution of this is p1 , and Eq. (2.41) is so simple that we can
obtain an explicit expression for this value here. It is
p1 = (140/3) psi pg1 = 32.0 psi
(however in the case of a substance whose equation of state is not so simple,

we must use the guessing and interpolation process described above). The final
temperature is, using the appropriate value for R
T2 = (186.67 psi)(3 ft3 /lbm)/(.6691 psi ft3 /lbm R)
which evaluates to 837 R (T f 2 = 377 ◦ F).
Equation (2.41) is simple enough to carry out the differentiations required to

obtain α and β. Doing this [see Example 2.2 and Eqs. (1.53) and (1.51)] gives
α = 1/T β = 1/ p (2.42)
Moreover, if you substitute these values into Eq. (2.30) and evaluate the integrals,
you will return to Eq. (2.41), as you must.
A polytropic change of state is one in which the properties of the initial and final
states are related by the equation
p1 v1n = p2 v2n = constant (2.43)
where n is a real number. We have already seen examples of this. For n = 0 the initial
and final pressures are equal, and for n = ∞ the specific volumes are the same. For
n = 1, the polytropic change of state together with Eq. (2.41) implies that the initial
and final temperatures are the same. The following problem shows how this sort of
information can be used to solve problems.
Example 2.10 One lbm of Neon is contained in a 4.5 ft3 cylinder at 90 psi
pressure by a piston that is pinned in place. What is the initial temperature?
After the piston is released, the system comes to rest with the piston 2.5 times
higher than it was initially. If the change of state is polytropic with n = 1.4,
what is the final temperature?
Solution. The problem information is summarized as follows:
State 1 State 2
M1 = p1 = p2 = (v1 /v2 )1.4 p1 M2 = M1
1 lbm 90 psi (polytropic) (fixed mass)
V1 = V2 = 2.5V1
4.5 ft3 (geometry)
The initial temperature is easily obtained by approximating the behavior of

Neon by Eq. (2.39) (R for Neon is 0.5316 psi ft3 /lbm R)

90 psi 4.5 ft3 /1 lbm 0.5316 psi ft3 /lbm R
T1 = p1 V1 /M1 R =
762 R
which corresponds to T f 1 = 302 ◦ F. The final temperature can also be expressed

by using Eq. (2.41). However when this equation is applicable, it is usually
computationally simpler to use the ratio of the two equations
p1 v1 /T1 = p2 v2 /T2 = R
because in this form R does not appear. Thus
T2 = ( p2 / p1 )(v2 /v1 )T1 = (v2 /v1 )1−1.4 T1
The final extensive conditions give V2 /V1 = v2 /v1 = 2.5 so that the final tem-
perature calculation looks like
T2 = (2.5)−0.4 762 R
the result of which is 528 R (T f 2 = 68 ◦ F).
Condensible Gases
Just as a liquid can vaporize and become a gas, a gas can condense and become a
liquid. Figure 2.8 is a sketch of the T, v plane of a condensible gas showing some
isobars along with the saturated vapor curve. The vapor specific volume, vg (T ), is
76 2 Equilibrium
Fig. 2.8 Contour plots of the state surface for a condensible gas showing the liquid and vapor
regions, and the saturated liquid and saturated vapor curves which coalesce at the critical point. In
the p, T plane both these curves appear as one, the saturation curve. A constant pressure curve is
sketched in the T, v plane, and two constant volume curves in the p, T plane
a decreasing function of temperature that coalesces with the saturated liquid curve,
described earlier, at the critical point. The saturated liquid curve and the saturated
vapor curve are two branches of the saturation curve for the substance. In the vicinity
of the saturated vapor curve, the equation of state cannot be accurately represented
by simple formulae. Therefore it is usual to simply tabulate this function, in the form
v = v( p, T ) p ≤ ps (T ) (2.44)
which we use directly when p and T are given as the independent variables. Notice
that when the vapor just begins to condense, p = ps (T ), this equation of state gives
v = v[ ps (T ), T ] = vg (T )
If the values of either p or T are not listed in the table, we approximate the state
surface by the nearest chord plane, Eq. (2.24), to numerically evaluate v. When T
and v ≥ vg (T ) are given, or when p and v ≥ vg [Ts ( p)] = v̂g ( p) are given, the chord
plane is solved and evaluated for the relevant variable.
The critical point is a singular point of the surface (see Fig. 2.6). From the graph
on the left of Fig. 2.8 you can infer that at the critical point ∂v T = 0 (at this point
∂v ∂v T = 0 also), so that ∂T v is not continuous there. Indeed at all the points of the
saturation curve the partial derivative ∂v T is not continuous. In the graphs in the
figure, you can see that on the saturation curve, v( p, T ), is double valued, which
physically just means that at a given pressure and temperature one can have a liquid
with specific volume, v f , and a vapor with specific volume, vg .
For water, the state surface corresponding to the gaseous phase (water vapor) is
given in a table of properties, and values for vg are listed together with the liquid
saturation curve data discussed earlier. The following problem illustrates the use
of these tables and indicates as well that we can solve problems involving varying
amounts of matter.
Example 2.11 Three lbm of water exist at 24 psi and 297 ◦ F in a rigid container.
A valve is opened and steam at 60 psi enters until the pressure inside equals
this value. At that point the valve is closed, and a thermometer indicates that
the temperature inside is 400 ◦ F. How much water entered the container?
Solution. We organize the information in our usual fashion
State 1 State 2
M1 = p1 = 24 psi p2 = 60 psi V2 = V1
3 lbm (stated) (rigid
Tf 1 = 297◦ F Tf 2 = 400◦ F container)
(stated)
The final specific volume can be read directly from the table, v2 = 8.354
ft3 /lbm, however, the final mass cannot be determined from this until the vol-
ume has been determined. This we obtain by considering state 1. Now the
values of p1 and T1 do not appear in the table. Therefore we need to use the
equation of the chord plane, Eq. (2.24), to interpolate for v1

v1 − v O = (T1 − TO ) + ( p1 − p O )
The point O is the one nearest to the state 1 values therefore p O is 20 psi and
TO is 280 ◦ F. Completing the calculation
⎧
⎨ 21.732 + [(10.708 − 21.732)/(40 − 20)](24 − 20)
v1 = +[(22.976 − 21.732)/(320 − 280)](297 − 280)
⎩
20.05 ft3 /lbm
The volume is therefore
V1 = v1 M1 = (20.05 ft3 /lbm)(3 lbm) = 60.17 ft3 ,
and the final mass is
M2 = V2 /v2 = V1 /v2 = (60.17 ft3 )/(8.354 ft3 /lbm) = 7.20 lbm
Thus
M2 − M1 = 7.20 lbm − 3 lbm = 4.20 lbm
is the amount of water that entered the container.

78 2 Equilibrium
2.4.3 Multicomponent Systems
Up till now we have considered equations of state that described systems containing
one phase of a single chemical constituent. However, many commonly encountered
systems are composed of more than one substance or more than one phase of a sub-
stance (or both). In order to solve problems involving these more complex systems,
we need to understand more about their equations of state.
The equation of state, in the form of Eq. (2.8), can be written in terms of the total
volume
V = Mv( p, T )
which we can recast as
V = M V ( p, T, M/M) = V ( p, T, M)
The last form results from the fact that mass and volume are extensive, while pressure
and temperature are intensive. Generalizing this result for a body composed of
simple compressible constituents, each of mass M j in macroscopic equilibrium at a
given pressure and temperature gives formally
V = V ( p, T, M1 , M2 , . . . , M )
The total mass of such a body is just the sum of the individual masses (recall
Sect. 1.3.2)

M= Mj
j=1
Since the constituent masses are all extensive properties, we can write the expression
for the volume as
V = M V ( p, T, M1 /M, M2 /M, . . . , M /M) = Mv( p, T, x1 , x2 , . . . , x−1 )

(2.45)
where x j is the mass fraction M j /M of the j th component. The state is specified by
knowledge of + 2 properties, of which one is extensive and + 1 are intensive.
For the case of a single constituent, = 1, Eq. (2.45) gives Eq. (2.8), whereas for
two constituents it becomes (note that x1 + x2 = 1)
v = V ( p, T, x1 , x2 ) = v( p, T, x1 ) (2.46)
This equation of state requires knowledge of three intensive properties in order to

determine the fourth, therefore it cannot be represented geometrically by a surface
in a three-dimensional space. Bodies composed of such combinations are capable
of very complex behavior, for example, they can undergo chemical reactions and
changes of phase. However if the amount of each constituent is fixed, the mass frac-
tions are simply parameters that do not affect volume changes. In such a case, the
subset of all equilibrium states corresponding to fixed values of the mass fractions is
a three dimensional surface, and such a material is effectively a simple compressible
substance. Air, which is composed of fixed amounts of gaseous Nitrogen, Oxygen,
and other elements is an example of such a substance. However at very low tempera-
ture when the elements begin to condense, this is no longer true; when condensation
of its elements occurs, air cannot be considered a simple compressible substance.
Example 2.12 Three lbm of air at 70 ◦ F is contained in a 6 ft3 part of an 18 ft3

vessel; the other part of the vessel is evacuated. When the dividing partition
is slid aside, the air expands freely to fill the entire vessel, and comes to rest
with a temperature of 70 ◦ F. At this final state, a pressure gauge connected to
the vessel reads 18 psi, and the atmospheric pressure is 14.7 psi. What was the
initial pressure, and what is the molar mass of air?
Solution.
State 1 State 2
M1 = Tf 1 = 70◦ F Tf 2 = 70◦ F V2 = 18 ft3
3 lbm
V1 = p2 = 32.7 psi M2 = M1
6 ft3 (fixed mass)
As I pointed out previously, it is worthwhile to use the ideal gas equation

of state in the form in which it combines information from both states. Doing
this here gives
p1 v1 /T1 = p2 v2 /T2
and everything is known except the initial pressure. Solving for p1 and using
the data to evaluate this produces
p1 = (v2 /v1 )(T1 /T2 ) p2 = 98.1 psi ( pg1 = 83.4 psi )
Actually more information is given here than is required, however, this allows
us to determine the molar mass through the gas constant
R = p2 V2 /M2 T2 = 53.34 ft lbf/lbm R
by means of Eq. (2.40)

m̄ = kU /R
In this way we obtain 28.96 lbm/lbmol for the molar mass of air. Of course
there are no molecules of air; 28.96 is some sort of average molecular weight,
Eq. (2.38).
80 2 Equilibrium
Fig. 2.9 Schematic showing

the variables that affect
phase equilibrium between a
liquid and its vapor
Phase Equilibrium
When water boils, the liquid phase and the vapor phase coexist with one another. We
depict this situation in Fig. 2.9 which shows a mass of liquid M f occupying a volume
V f in contact with a mass of vapor Mg occupying a volume Vg . These two systems
are in equilibrium with each other and therefore their temperatures and pressures are
equal; the combination has an equation of state like Eq. (2.46). Moreover the common
temperature is the saturation temperature, and therefore the common pressure must
be the saturation pressure, p = ps (T ). This means that in Eq. (2.46) we have
v = v[ ps (T ), T, x]
which depends only on two intensive variables, and therefore represents a simple
compressible substance.
As we discussed in previous sections, the intensive variables v f and vg are also
functions only of the saturation temperature (or saturation pressure), so that knowl-
edge of only one of the four quantities, ps , Ts , v f , vg , determines them all. Of the four
extensive variables, V f , M f , Vg , Mg , the volume and mass of the liquid and vapor,
respectively, knowledge of one liquid and one vapor variable suffices to determine
the others through the relations
Vf = Mf vf Vg = Mg vg
Therefore, this equilibrium is completely specified by the knowledge of one intensive

and two extensive variables, and we can regard it as a state of a simple compressible
substance.
The equation of state is obtained by using the extensive character of the total
volume and total mass
V = V f + Vg M = M f + Mg
Writing the first of these as

V = M f v f + Mg vg
dividing by the total mass, and rearranging, gives the explicit form of Eq. (2.46)
applicable to this case
v = v f + x(vg − v f ) (2.47)
where x is the mass fraction of the vapor Mg /M; it is called the quality of the steam.
In states involving liquid–vapor equilibrium there are 12 variables,
V, M, V f , M f , Vg , Mg , v, p, T, x, v f , vg
however they are related by 9 equations so that 3 of the variables specify the
rest. Recall that p, T, v f , vg cannot be specified together since they are dependent.
Some other dependent combinations are also prohibited, for example, V, M, v and
M, Mg , x. All this can be summarized in tabular form.
State Properties Equations
M p Liquid Vapor Substance
V T vf = vf (T ) vg = vg (T ) p = ps (T )
Mg vg Vf = Mf vf Vg = Mg vg V = Mv
Vg vf v = vf + x(vg − vf )
Mf v x = Mg /M
Vf x M = Mf + Mg
In most thermodynamics problems we have little interest in the mass and volume of
the individual phases Mg , M f , Vg , V f . This leaves 8 properties to determine, instead
of 5 for a pure substance. The following example illustrates these points.
Example 2.13 Three lbm of water is contained in a 2 ft3 rigid tank at a temper-
ature of 240 ◦ F. What is the quality of the steam? What fraction of the volume
is vapor? Vapor is withdrawn while the temperature remains the same until the
water exists as saturated vapor. How much mass was withdrawn?
Solution.
State 1 State 2
M1 = Tf 1 = T2 = T1 V2 = V1
3 lbm 240◦ F (stated) (rigid tank)
V1 = x2 = 1
2 ft3 (saturated vapor)
The specific volume in state 1 is V1 /M1 = 0.6667 ft3 /lbm, and at the
given temperature, the specific volume of the liquid and vapor phases are
0.016926 ft3 /lbm and 16.321 ft3 /lbm, respectively. Therefore the quality is,
from Eq. (2.47)
x1 = (v1 − v f 1 )/(vg1 − v f 1 ) = (0.6667 − 0.016926)/(16.321 − 0.016926) = 0.0398

82 2 Equilibrium
This means that 3.98% of the mass of water initially is vapor. The volume of
vapor is obtained from
Vg1 = vg1 Mg1 = vg1 x1 M1
or numerically
Vg1 = (16.321 ft3 /lbm)(0.0398 · 3 lbm) = 1.95 ft3
The volume fraction is Vg1 /V1 or 0.976, which means that almost 98% of the
volume is vapor even though it is only about 4% of the mass. Of course this
happens because the liquid is much more dense than the vapor.
The final specific volume is just vg at 240 ◦ F which we have already found
as 16.321 ft3 /lbm. Therefore the final mass is
M2 = V2 /v2 = (2 ft3 )/(16.321 ft3 /lbm) = 0.0408 lbm
and the amount of water withdrawn was 2.96 lbm.
The liquid, Eq. (2.34), liquid–vapor, Eq. (2.47), and vapor, Eq. (2.44) equations
of state of a substance can be combined to form the state surface. This surface for
water is illustrated in Fig. 2.10, which also includes the solid state and the solid–
liquid, and solid–vapor equilibrium states. We have also sketched in Fig. 2.10, the
contour representation of this surface in the p, v plane (here you can infer that at
the critical point ∂v p = ∂v ∂v p = 0). You can see the other contour planes of this
surface in Fig. 2.8. These include the saturation curve as an aid in locating the state
of the substance relative to its phase. It is often helpful to sketch the equilibrium state
points in these planes when solving thermodynamics problems.
Example 2.14 Saturated liquid refrigerant-12 at 20 ◦ F is heated in a piston

cylinder until the piston reaches a set of stops located at a height 11.5 times
greater than the initial height. If the heating continues until the liquid has
completely vaporized, what is the final temperature?
Solution.
State 1 State 2
Tf 1 = 20◦ F x2 = 1 M2 = M1
x1 = 0 (completely (fixed mass)
(Saturated vaporized) V2 = 11.5V1
Liquid) (geometry)
Fig. 2.10 The state surface for a substance that has solid, liquid, and vapor phases, and like water,
expands upon freezing. We also show the p, v plane with some isotherms, and the corresponding
space curves on the surface
Since only five properties are given, we cannot determine all the properties.
We can determine all the intensive properties but we cannot know the masses or
the volumes. The final temperature is the saturation temperature corresponding
to the vapor specific volume. This is, in implicit form
v2 = vg2 = vg (T2 )
The extensive equations of state 2 provide, as we have seen before, a relation-

ship between the specific volumes in the two states
v2 = 11.5v1
Now v1 is just the saturated liquid specific volume at 20 ◦ F, 0.011307

ft /lbm. This gives v2 as 0.130030 ft3 /lbm. The change of state is sketched on
3
the accompanying T, v plane, and this may help you visualize what is going on.
The dotted line signifies a hypothetical path between the states. With v2 known,
the implicit equation above can be solved by approximating the saturated
84 2 Equilibrium
vapor curve by a chord line about the point. In this way we find
T f 2 = 160 + [(180 − 160)/(0.10499 − 0.13804)](0.13003 − 0.13804)
which gives the final Fahrenheit temperature as 165 ◦ F.
By analogy with static equilibrium in which the pressure in the two phases are
equal, and thermal equilibrium in which the temperature in the two phases are equal,
we imagine that there is a property in each phase whose equality corresponds to
material equilibrium. Denoting this property by g, we have
g f ( p, T ) = gg ( p, T ) (2.48)
The equation of the saturation curve
p = ps (T )
that we have discussed previously is the result of solving Eq. (2.48) for the saturation
pressure in terms of the saturation temperature. Although we have no idea at this
point how to express g, we will discover this later in our study.
By generalizing the argument to a situation in which components coexist in P
phases, it can be shown that only
D =− P +2 (2.49)
intensive variables are independent; the number D is called the number of degrees of
freedom. This equation is called Gibbs phase rule, in honor of its discoverer, J. Willard
Gibbs, who was the first American physicist to achieve international recognition
for his scientific work. He obtained the result by observing that an equation like
Eq. (2.48) exists for every pair of phases. There are, in addition, P independent
extensive variables, for a total D + P properties that specify the state. In the case
when D + P = 3, the material can be regarded as a simple compressible substance.
As we have seen for the two phases of water, D = 1; once the temperature is set, the
pressure must be the saturation pressure in order for equilibrium to be maintained.
When three phases of water coexist the phase rule produces D = 0; no intensive
variable can be set arbitrarily, in other words, there is a unique temperature and
pressure for this triple point. In this case three extensive variables, the masses of the
water in each of the three phases completely specify the state, which can also be
regarded as that of a simple compressible substance.
2.5 Exercises 85
2.5 Exercises
In each of the following exercises where appropriate, indicate the change of state on
a suitable projection plane of the state surface. For substances that can undergo a
phase change sketch the saturation curve as well in this diagram.
Section 2.2
2.1 Use the expression v = RT / p to integrate Eq. (2.3), with the initial condition
p = pa at z = 0, for the pressure distribution in an equilibrium atmosphere. Evaluate
the characteristic height h ∗ = RT /g for a temperature of 50 ◦ F (use R = 53.34 ft
lbf/lbm R). Plot the dimensionless pressure p/ pa against the dimensionless altitude
z/ h ∗ (in the range 0 ≤ z/ h ∗ ≤ 3) using any convenient graphics software. Make
a Taylor expansion of the exponential keeping two terms (z << h ∗ ); the result is
Eq. (2.5).
Section 2.3.3
2.2 A gas is contained in a spherical balloon of radius 6 in at 15 psi and 70 ◦ F. The
balloon is filled until the pressure is 30 psi. At that point the radius is 10 in while the
temperature is the same. How much mass was added to the balloon?
2.3 A substance of mass M is contained in a rigid volume V at temperature T ∗ . If

half the mass is released so the pressure drops to 70% of the initial pressure, what is
the final temperature?
Section 2.4.1
2.4 Three lbm of Ethyl alcohol is contained in a rigid cylinder at 70 ◦ F and 14.7
psi. If the cylinder is heated until it is 90 ◦ F, what is the final pressure? (1600 psi)
2.5 One kg of glycerin is stirred in a vat that is open to the atmosphere. The process
continues until the final temperature is 50 ◦ C higher than it was initially. What is the
change in volume? Hint: There is not enough information given to determine both
volumes, however, the difference in volumes can be found.
2.6 Water is contained in a 3 ft3 rigid tank at 450 psi and 340 ◦ F. A valve is opened
and liquid is released until the remaining water just begins to boil at a pressure of
235 psi. What mass of water was released? (6.3 lbm)
2.7 One kg water, contained in a rigid tank at 100 ◦ C, 30 MPa is cooled until the
water just begins to boil. What is the final temperature, and pressure, and what is the
volume of the tank?
Section 2.4.2
2.8 For an ideal gas, p = RT /v, derive the results for α and β given in Eq. (2.42).
86 2 Equilibrium
2.9 For a van der Waals gas, p = RT /(v − b) − a/v 2 , where a and b are constants,
show that α and β are given by

RT v 2a(v − b)
α = R/ −
(v − b) v2

RT v 2a
β = 1/ − 2
(v − b)2 v
To do this most easily use the appropriate form of Eq. (2.26), one of which is
Eq. (2.28). This trick is needed because you cannot easily solve the van der Waals
equation explicitly for v.
2.10 Set ∂v p = 0 in the van der Walls gas of Exercise 2.9. This gives a curve
T = T (v) that separates the region where ∂v p > 0 from the one where ∂v p < 0
in the T − v plane. Substitute the equation itself into the equation to obtain the
separation curve in the p − v plane, p = p(v). Sketch these curves and show by
differentiation that the maximum occurs at v = vc = 3b, T = Tc = 8a/(27Rb), and
p = pc = a/(27b2 ). The maximum corresponds to the critical point and while the
curves are not saturation curves they are closely related. The area under the curves is
not physical, ∂v p > 0. The equation is important because it models gas condensation
due to finite particle volume, b, and inter-particle forces, a/v 2 .
2.11 Calculate the values of α and β for steam at 100 psi and 400 ◦ F using infor-
mation found in the tables.
2.12 Calculate the values of α and β for R-12 at 0.6 MPa and 60 ◦ C using information
found in the tables.
2.13 Show that using Eq. (2.42) for α and β in the equation of state Eq. (2.30), and
integrating, you obtain the ideal gas equation of state.
2.14 One and a half lbm of Argon is heated in a piston cylinder arrangement from
30 psi and 300 ◦ F until the height increases by 30%. If the final pressure is the same
as initially, what is the final temperature?
2.15 One and a half lbm of Argon is heated in a piston cylinder arrangement from 30
psi and 300 ◦ F until the height increases by 30%. If the change of state is polytropic
with exponent −1.2, what is the final temperature, and pressure? (893 ◦ F, 41.1 psi)
2.16 Helium is contained in a spherical balloon of radius 6 in at 15 psi and 70 ◦ F.

The balloon is filled until the pressure is 30 psi. At that point the radius is 10 in while
the temperature is the same. What mass of Helium was added to the balloon?
2.17 Two lbm of steam is cooled in a piston cylinder arrangement from 220 psi,
530 ◦ F until it just begins to condense at 200 psi. If the piston diameter is 10 in, what
is its change in height, z 1 − z 2 ? (0.991 ft)
2.5 Exercises 87
2.18 One lbm of R-12 is contained in a piston cylinder arrangement at 20 psi and
a volume of 1.7 ft3 . The gas is heated and the piston rises until it encounters stops
when the volume is 1.9 ft3 , however, the heating is continued until the pressure has
risen to 30 psi. What is the final temperature?
2.19 Steam is contained in a 1.7 ft3 tank at 600 psi and 500 ◦ F. A valve is opened and
the steam expands into a larger volume until its pressure is 110 psi and its temperature
is 360 ◦ F. What is the volume of the additional space into which the steam expanded?
(7.40 ft3 )
2.20 Three kg of Neon is contained in a cylinder at 50 ◦ C by a steel piston 0.8

m in diameter and 30 cm thick. The gas is heated and the piston begins to rise. It
immediately contacts a linear spring with constant 5 N/mm which is compressed in
the process. When the heating ceases and equilibrium is reestablished the piston is
10% higher than it was originally, what is the temperature? Atmospheric pressure is
101 kPa.
2.21 Three kg of Neon is contained in a cylinder at 50 ◦ C by a steel piston 0.8

m in diameter and 30 cm thick. The gas is heated and the piston begins to rise. It
immediately contacts a linear spring with constant 5 N/mm which is compressed in
the process. When the heating ceases and equilibrium is reestablished the temperature
is 300 ◦ C, what is the pressure? Atmospheric pressure is 101 kPa. (153 kPa)
2.22 Three kg of steam is contained in a 0.8 m diameter cylinder at 200 ◦ C and 0.8
MPa. The vapor is heated and the piston begins to rise. It immediately contacts a
linear spring with constant 50 N/mm which is compressed in the process. The heating
ceases and equilibrium is reestablished when the temperature is 400 ◦ C. At that point
what is the pressure?
2.23 One kg of air is contained in a 3 m3 container at 1 MPa. A valve is opened

and the air expands into a volume 3 times the original. If the final temperature is the
same as the initial temperature, what is the final pressure? (250 kPa)
2.24 Saturated water vapor at 200 ◦ C is contained in a cylinder fitted with a piston
whose volume is 25 l. If the steam expands until the pressure is 200kPa at the same
temperature, what is the final volume? Consider this to be a polytropic change of
state, and determine the exponent n.
2.25 One-quarter kg of air is contained in a cylinder at 100 kPa and 30 ◦ C. The

piston is impacted by a force that compresses the air to 1/10 its original volume in
a change of state that is polytropic with exponent 1.35. What are the final pressure
and temperature? (2.24 MPa, 405 ◦ C)
2.26 A rigid cylinder containing 1.5 kg of O2 at 125 kPa, 27 ◦ C, and another rigid
cylinder containing 2 kg of O2 at 147 kPa, 35 ◦ C are connected by a pipe with a
closed valve. If the valve is opened, the contents of the two cylinders mix, and the
final common temperature is 30 ◦ C, what is the final common pressure?
88 2 Equilibrium
Section 2.4.3
2.27 Two lbm of steam is contained in an 8 ft3 rigid vessel at 450 ◦ F. The steam is
cooled until the temperature is 240 ◦ F. What are the initial and final pressures, and
what is the final fraction (by mass) of vapor? (131.2 psi, 24.968 psi, 24.4%)
2.28 One kg of R-12 is cooled in a cylinder from 1.6 MPa, 100 ◦ C to 1/5 its initial
volume. If the final temperature is 40 ◦ C, what is the final quality?
2.29 R-12 is contained in a vertical piston cylinder device which has a linear spring
fitted to a 6 in diameter piston. Initially, the volume occupied by the R-12 is 1.6 ft3 ,
the temperature is 40 ◦ F, and the quality is 30%. The system is heated until the R-12
exists as saturated vapor at 94 ◦ F. What is the value of the spring constant and how
much mass is contained? (48.7 lbf/in, 6.62 lbm)
2.30 R-12 is contained in a vertical piston cylinder device which has a linear spring
with a constant of 50 lbf/in fitted to a 6 in diameter piston. Initially the volume
occupied by the R-12 is 1.6 ft3 , the temperature is 40 ◦ F, and the quality is 30%. The
system is heated until the temperature is 140 ◦ F. What is the final pressure?
2.31 One lbm of R-12 is contained in a 1 ft3 rigid tank at 30 psi. What fraction
of the substance (by volume) is vapor? If a valve is opened and vapor escapes until
all the remaining substance is vapor contained at the initial temperature, how much
R-12 was released? (0.233 lbm)
2.32 A vertical cylinder fitted with a piston contains 30 l of R-12 at 20 ◦ C, 90%

quality. The piston has a mass of 90 kg, a cross-sectional area of 0.006 m2 , and is
pinned in place. What is the mass of R-12 contained in the cylinder? If the atmospheric
pressure is 100 kPa, and the pin is removed so that the piston moves to a new
equilibrium with the R-12 temperature the same as it was originally, what is the final
volume?
Chapter 3
Work and Heat
3.1 Introduction
In the last chapter, we saw that changes of a macroscopic equilibrium state can be
accompanied by, or even result from, doing work on or heating a thermodynamic
system. Indeed, the ability to calculate these quantities is of the utmost importance to
engineers who need to know, for example, how much work must be provided in order
to supply compressed air at some required design pressure. Therefore, this chapter
is devoted to an extended discussion of each of them.
3.2 Mechanics
As we saw in Sect. 1.3.7, when the point of application of a force, F(x, v, t), moves
along a curve, x = xC (t) with velocity, v = vC (t), from time t1 to a later time t2 , the
force does an amount of work defined by
t2
WF (t1 , t2 , C) = F[xC (t), vC (t), t] · vC (t) dt (3.1)
t1
WF is an interaction, its value depends on C (see Sect. 2.3.2). The integrand here is
the rate at which work is done by the force; we denote it by
ẆF (x, v, t) = F(x, v, t) · v (3.2)
The quantity ẆF is a derived dimension [E/t] = [FL/t] = [ML2 /t3 ] (the dot over
the symbol is meant to indicate a rate, and WF is a derived dimension [E]), we call it
the power developed by the force. When the power developed is positive, we say that
the force generates power, and when it is negative, we say that it absorbs power. The
natural units for power are kg m2 /s3 in the SI, and lbm ft2 /s3 in the English System.

90 3 Work and Heat
The first of these is called a watt (W) and is routinely used, however, the second has
no name and is not used. Instead, as usual in the English System, we use a subsidiary
unit, the horsepower (hp), defined (by James Watt) as
1 hp = 550 ft lbf/sec = 1770 lbm ft2 /sec3
The average power developed is given by (this is the usual definition of an average,
see Sect. 1.3.8, Kinetic Model for Pressure)
t2
1 WF (t1 , t2 , C)
{ẆF } = ẆF dt = (3.3)
t2 − t1 t1 t2 − t1
This is a useful quantity for rating and comparing motors.
3.2.1 Conservative Forces
In the situation usually studied in elementary mechanics, the moving force depends
neither on velocity nor explicitly on time, and the integral above is parameterized as
a line integral in space (see Sects. 2.3.1 and 2.3.2)

WF (x1 , x2 , L) = F(x) · d x (3.4)
L
Although WF , in general, depends on the path L that connects the initial, x1 , and final,
x2 , points, there is a special class of forces, called conservative forces, for which the
work done is independent of this space curve. Consequently, conservative forces can
be moved any distance along any path and returned to their starting point along any
other path, with no network having been done. All the work required to move the
force away from the starting point is recovered on the return.
In two dimensions, (y, z), for which the force has components (Fy , Fz ), Eq. (3.4)
can be written as [see also Eq. (2.21)]
y2 z2
WF = Fy [y, zL (y)] dy + Fz [yL (z), z] dz
y1 z1
A two-dimensional conservative force is defined by [see Eqs. (2.19) and (2.23)]
∂z Fy = −∂z ∂y φF = −∂y ∂z φF = ∂y Fz
so for such a force, the line integral is independent of the path, L [see Eq. (2.23)],
and
WF (x1 , x2 , L) = φF (y1 , z1 ) − φF (y2 , z2 )
3.2 Mechanics 91
The state function φF is called the potential energy. It is a relative quantity, however,
the work done is an absolute quantity (see Sect. 1.2.2). The best known conservative
force is the weight of a mass, m, for which
WG (z1 , z2 ) = mgz1 − mgz2 (3.5)
Here the relative aspect of the gravitational potential energy, φG = mgz, is reflected
in the fact that the height, z, can be measured from an arbitrary level.
3.2.2 Reversible and Irreversible Work
A characteristic of all forces (nonconservative as well as conservative) that are inde-

pendent of velocity and time, is that the work done by them is reversible. This means
that if WF (t2 , t1 , C) is the work done by F in traversing C in reverse, then
WF (t1 , t2 , C) + WF (t2 , t1 , C) = 0
In other words, such a force can be moved any distance along any path C, and returned
to its starting point along the same path, with no network having been done. All the
work required to move the force away from the starting point is recovered on the
return.
In thermodynamics we are usually interested in forces that do not depend on time
explicitly. The simplest way to determine whether the work done by such forces,
F(x, v), is reversible or not is to use Eq. (3.2) and to substitute −v for v (this corre-
sponds to traversing C in reverse). If
ẆF (x, v) = −ẆF (x, −v) (3.6)
the work done by the force is reversible. As I indicated at the outset of this section,
any force which is independent of both velocity and time satisfies Eq. (3.6); you can
verify this by using Eq. (3.2).
Not all forces do work reversibly. If
ẆF (x, v) = ẆF (x, −v) (3.7)
we say that the work done by the force is irreversible. In general, the work done by
a force has both reversible and irreversible components
ẆF = ẆFR + ẆFI
because an arbitrary function can be written as the sum of odd and even functions
[Eq. (3.6) is the mathematical definition of an odd function of v while Eq. (3.7) is
the definition of an even function].
92 3 Work and Heat
Friction is the most common type of force for which the work done is irreversible.
The reason for this is that a friction force always opposes the velocity of its point of
application. For example, the usual expression for Coulomb friction1
Ff = −μN v/|v|
when substituted into Eq. (3.2) gives
ẆFf (x, v) = ẆFf (x, −v) = −μN |v|
Here it is easy to see that Coulomb friction requires work to move it in any direction,
because the power developed is always negative. Experience shows that this is true
of all types of friction. We say that friction is dissipative, since a friction force always
absorbs power when it moves.
3.2.3 Continuous Systems
A closed thermodynamic system, regarded as a continuum, exerts forces on its sur-

roundings across its enclosing surface. When we speak of the work done by a ther-
modynamic system, S, we mean the work done by these contact forces when the
system boundary moves. For a change of state that takes place from time t1 to a later
time t2 we write this as t2
WS = ẆS dt (3.8)
t1
Here, I have not written the arguments of work explicitly. In the interest of simplicity
I will continue this practice, however, it is important for you to remember that work
is not a property of a system, it is motion dependent. Moreover, when there can be
no confusion about which system is being considered, I will eliminate the subscript
denoting the system, and simply write W for the work done. When the work done by
the system is negative, we say that work is done on the system by the surroundings.
If we subdivide the surface of a system into N elements, see Fig. 3.1, then at
some instant during a change of state in which the j th element moves with velocity,
vj , the rate at which the system does work on the external force, Fj , that acts on the
element, is given by
−Fj · vj
Adding all these together gives the rate at which the system does work,
1 Hereμ is the coefficient of friction, and N is the component of force normal to the direction of
motion.
3.2 Mechanics 93
Fig. 3.1 The jth element of

the surface of a system,
along with its normal, the
force exerted on it by its
surroundings, and its velocity
N
N

Ẇ = −Fj · vj = −τ nj · vj Aj (3.9)
j=1 j=1
The second equality follows from the definition of traction in Sect. 1.3.8. On subdi-
viding indefinitely, the finite sum becomes a surface integral

Ẇ = −τ n · v dA (3.10)
A
Conservation of Power
Imagine a system, S, in contact with a number, N of other systems, Sj , as I have

indicated schematically in Fig. 3.2. The surface, AS , of S can be divided into portions,
which we denote by AS/Sj , that separate S from each of the surrounding systems Sj ,
and since integration is an additive operation, the power developed by S is just the
sum
ẆS = ẆS/S1 + ẆS/S2 + · · · + ẆS/SN (3.11)
where the rate of doing work by S on Sj is from Eq. (3.10)

ẆS/Sj = −τ n · v dA
AS/Sj
Fig. 3.2 A system in contact

with several other systems.
The outer surface of the
collection, which is the
darker curve, does not move;
the collection is a
mechanically isolated system
94 3 Work and Heat
The rate of doing work by Sj on S is analogously

ẆSj /S = −τ −n · v dA
AS/Sj
Here we are assuming that both bodies have the same velocity, v, at their common
surface; namely, that there is no slipping between them. Then, since Newton’s third
law, action equals reaction, is
τ −n = −τ n
we find that
ẆS/Sj + ẆSj /S = 0 (3.12)
This is the law of conservation of power, which means physically that provided there
is no slipping, the power generated by S in moving Sj , is equal to the power absorbed
by Sj when moved by S. Integrating over time, we obtain a similar result for the work
done
WS/Sj + WSj /S = 0 (3.13)
Although these results may seem trivial to you, some related, and very useful concepts
may not. For example, if the system U , composed of S and all the Sj as shown in
Fig. 3.2, is mechanically isolated, meaning that its boundary does not move, then
from Eq. (3.10), ẆU = 0 (since v = 0 on A). Moreover Eqs. (3.11) and (3.12) can
be used to show that
ẆU = ẆS + ẆS1 + · · · + ẆSN = 0 (3.14)
or in integrated form for the work done
WU = WS + WS1 + · · · + WSN = 0 (3.15)
These results mean that in a mechanically isolated system all the power generated in
one place is transferred to other places where it is absorbed.
3.2.4 External Determination of Work Done
Although Eq. (3.10) is appropriate for general motions of a deformable substance and
you will use it in fluid mechanics, in our study of the fundamentals of thermodynamics
we are usually interested in two special types of surface motion; either translations,
which occur when every point on the surface has the same velocity, or rotations,
for which the speed of every surface point depends linearly on its distance from a
fixed axis. In each case, we use the development of the previous section to relate the
3.2 Mechanics 95
work done by a system to that of its surroundings, and in that way determine it more
simply than by directly evaluating Eq. (3.10).
Rigid Body Translation
In this case every moving point on the surface of the system, S, has the same rectilinear
velocity, v = vi, so that from Eq. (3.10) we obtain
ẆS = FS/P · v (3.16)
where FS/P is the resultant force exerted by the system (note the change in sign) on
the moving element P.
This situation is commonly encountered in thermodynamics, a number of devices
that operate this way are depicted in Figs. 3.3 and 3.4. In each case, we relate the
force exerted by the system to the forces exerted by the surrounding systems through
the Newton’s law that applies to the translating element P
Fig. 3.3 Two examples of translational rigid body motion are the shock absorber sketched on the
left, and the shaker on the right. Both these devices undergo cyclic processes in normal operation
Fig. 3.4 Two additional examples of translational rigid body motion are the piston cylinder sketched
on the left, and the mixer/expander on the right. Both of these devices normally operate in the change
of state mode
96 3 Work and Heat
d 2x
mP i ≡ mP a = FS/P + GP + FO + FL/P + FA/P
dt 2
Here mP is the piston mass, FO , FL/P , and FA/P are the forces exerted by the out-
put/input device, the lubricant, and the air, respectively, on the moving element P,
and GP is the gravitational force acting on P. Solving explicitly for the force FS/P
gives
FS/P = −FO − FL/P − FA/P − (GP − mP a) (3.17)
and incorporating this expression into Eq. (3.16) results in
ẆS = [−FO − FL/P − FA/P − (GP − mP a)] · v (3.18)
On using the definition of power developed for the various systems, we see that this
is just a special case of Eq. (3.14)
ẆS = −ẆO − ẆL − ẆA − ẆP
which we condense by calling S̃ the surroundings of S and writing
ẆS̃ = ẆO + ẆL + ẆA + ẆP
so that on integration we get
WS = −WS̃ = −WS̃R − WS̃I (3.19)
In thermodynamics, we are principally interested in the work done by a translating

system S when it undergoes one of two types of process:
1. a cyclic process in which there is a reciprocating motion of P such that x, v, and
all force components acting on P are periodic functions with period tc , the time
to complete one cycle.
2. a change of state process in which S is in an equilibrium state at time t1 and is in
a different equilibrium state at t2 ,
Of the four terms in Eq. (3.18), we can integrate, at least partially, two of them in
general, and evaluate them for each type of process.
For example, the work done by P, WP , is obtained, on integrating the last term in
Eq. (3.18), as [see Eqs. (3.5) and (1.34)]
WP = −[mP g(z2 − z1 ) − mP (v22 − v12 )/2]
where z is the height of a point on P. Therefore in cyclic processes, for which v1 = v2

and z1 = z2 , and change of state processes, for which v1 = v2 = 0, and z1 = z2
(horizontal motion), we see that WP = 0. However, in change of state processes for
which P moves vertically we have
3.2 Mechanics 97
WP = −mP g(z2 − z1 )
Applied to a piston cylinder device with piston area AP , this can be written in terms
of the thermodynamic variable volume (see Fig. 3.4 and take z = x)
WP = −(mP g/AP )(V2 − V1 ) (3.20)
The power developed by the air, ẆA , can be written in a form that permits us to
obtain a general integral for an important part of it. Note first that we can write the
force exerted by the air on P in two parts [using Eq. (1.35)]

FA/P = − pa nA dA + (τ nA + pa nA ) dA
AA/P AA/P
where nA is the outward normal to A. The first part is due to the static air pressure,
and the second part is due to the motion. The corresponding work done by the air
also has two parts and can be obtained from Eq. (3.10) as

ẆA = ẆA/P = pa nA · v dA + (−τ nA − pa nA ) · v dA
AA/P AA/P
The second integral arises as a result of friction, consequently it is dissipative and

contributes only to the irreversible component of ẆA , as is the case with all types
of friction. The first integral, with pa removed since it is a constant, can be cast in
a very simple form, which is most easily understood by following the development
illustrated in Fig. 3.5. This leads to
N
V2 − V1
Aj nj · vj ∼
j=1
t2 − t1
Fig. 3.5 The dashed curve is the boundary of S at t1 , and the solid curve is the boundary at t2 ,
therefore the shaded area denotes the change in volume. This is composed of the sum of strips like
the one shown here in darker shading, whose approximate volume is given by the analysis in the
figure
98 3 Work and Heat
which, in the limit of an infinite number of surface elements and infinitesimal time
increment is exact,
dV
n · v dA = (3.21)
A dt
Putting −v for v here (by multiplying both sides of the equation by −1) reverses
the sign on the right hand side. Therefore, the first term of ẆA is reversible [recall
Sect. 3.2.2, Eq. (3.6)] and we write
d VA dV
ẆA = pa + ẆAI = −pa + ẆAI (3.22)
dt dt
in which VA is the volume of the surrounding air, and V is the volume of the system
S (since P translates, an increase in VA must result from a decrease in V ).
From this development, it is clear that whenever there is no volume change, there
is no reversible power developed by the air, and that in each cycle of a cyclic process,
for which V2 = V1 , zero work is produced reversibly. Moreover the work done by
the air, WA , in a change of state process is obtained by integrating Eq. (3.22)
WA = −pa (V2 − V1 ) + WAI (3.23)
Combining these results for cyclic process devices gives
ẆS̃ = ẆO + (ẆL + ẆAI ) = FO · v + (ẆL + ẆAI )
from which we obtain WS by integrating and substituting into Eq. (3.19).

The terms in parentheses here are irreversible and for shock absorbers are negligi-
ble compared with WO . Therefore, we can calculate the work done by these devices
in terms of the motion, v, and the force exerted, FO .
Example 3.1 A shock absorber is forced sinusoidally with a maximum force

of 100 lbf and at a frequency of 50 Hz. If the motion is also sinusoidal, lags
the force by 90◦ , and has a stroke of 4 in, how much work is done by the shock
in each cycle of operation? What is the average power dissipated?
Solution.
According to the problem statement we can write
FO = F0 sin ωt i x = x0 sin(ωt − π/2) i = −x0 cos ωt i
where the stroke is 2x0 , the maximum force is F0 , and the frequency is related
to the cycle time f / cycle = 1/tc , and angular frequency ω = 2πf / cycle. The
velocity of motion is obtained by differentiating the position vector
v = x0 ω sin ωt i
3.2 Mechanics 99
so the work done by the surroundings in one cycle is

tc
WS̃ = F0 x0 ω sin2 ωt dt = F0 x0 ωtc /2 = πF0 x0
0
accordingly the work done by the shock in one cycle is
WS = −WS̃ = −3.14159 · 100lbf · 2in = −628.3in lbf
The minus sign means that it requires work to operate the shock. The average
power developed is given by Eq. (3.3), {ẆS } = WS /tc , so that

−WS f / cycle
−{ẆS } =
(628.3in lbf/cycle · 50cycle/s)/(12in/ft · 550ft lbf/hp s)
which is 4.76 hp. Note that −{ẆS }, is the negative of the power developed,
and so indicates the power dissipated.
When a motor of known power output is connected to a shock absorber, and

operated for a specified length of time, tO , the work done on S is simply
WS̃ = WS̃R = {ẆO }tO = {ẆM }tO = WM
where WM is the work done by the motor in the time tO . As indicated in Fig. 3.6, in
a cyclic process, the instantaneous power developed, ẆS , is a time varying periodic
function whose average is {ẆS }.
When the irreversible component of work done, WS̃I is not negligible, we need to
specify additional information to determine WS . This information is usually given in
terms of a positive ratio, ηM , called the mechanical efficiency
Fig. 3.6 The area under the

instantaneous power curve
over a cycle, WS (dark
shading), is equal to {ẆS }tc
(light shading) as given by
Eq. (3.3)
100 3 Work and Heat

1 + WS̃I /WS̃R ≤ 1 WS̃> 0
ηM = R
(3.24)
1/[1 + WS̃I /WS̃R ] ≤ 1 WS̃<
R
0
It makes sense to call this ratio an efficiency because 0 ≤ ηM ≤ 1. You can see
this from Eq. (3.19) by recalling that the irreversible component of work is always
negative, WS̃I ≤ 0. When we know both WS̃R and ηM , either from measurements or
calculations we use this definition and Eq. (3.19) to obtain WS

−WS̃R /ηM WS̃R < 0
WS = (3.25)
−ηM WS̃R WS̃R > 0
The irreversible component, WS̃I , is composed of the work done by the lubricant,
which we call the bearing loss, and the irreversible component of the work done by
the air, which we call the windage loss, thus
WS̃I = WL + WAI
which we determine either from Eq. (3.19), WS̃I = −WS − WS̃R , or by solving for it
in Eq. (3.24)
(1 − ηM )WS̃R /ηM WS̃R < 0
WS̃I =
(ηM − 1)WS̃R WS̃R > 0
We can use all this as follows.
Example 3.2 A 3/8 kW motor operates a shaker containing a substance to be

mixed. If the motor runs for 12 minutes, and the dissipation is 43 kJ, how
much work is done on the substance in the shaker and what is the mechanical
efficiency?
Solution.
According to Eq. (3.3) the reversible work done by the surroundings is
WS̃R = {ẆM }tO = 0.375kJ/s · 12min · 60s/min = 270 kJ
Therefore the work done by the substance is from Eq. (3.19)
WS = −WS̃R − WS̃I = −270 kJ − (−43 kJ) = −227 kJ
and the efficiency is obtained from Eq. (3.24) as
ηM = 1 + (−43)/270
which is 84%.
3.2 Mechanics 101
Just as we did for cyclic processes, for processes involving a change of state, we
can calculate (in simple cases), or measure (in more realistic ones) the components
of WS̃ and thereby obtain WS , from Eq. (3.19). For example, if a piston moves a
constant load L [see Eq. (3.20)]
WO = −(L/AP )(V2 − V1 )
while if the system connected to the piston were a linear spring with constant k and
free length 0 , then we would have
WO = −(k/2)[(2 − 0 )2 − (1 − 0 )2 ]
or on relating the spring length to system volume
WO = −(k/2A2P )[(V2 − V0 )2 − (V1 − V0 )2 ] (3.26)
where V0 is the cylinder volume when the spring exerts no force.

The reversible component of the work done by the surroundings,
WS̃R = −pa (V2 − V1 ) + WP + WO
can be either positive or negative in this case. In other words, the surroundings can
either do work or have work done on them. Note that WP is either zero for horizontal
motion or given by Eq. (3.20) for vertical motion. The following example illustrates
these concepts.
Example 3.3 Three kg of air is contained in a piston cylinder device at 400 kPa
and 25 ◦ C. The piston carries a static load, including its own mass, of 20 kN,
but is otherwise unconnected to an external system. The air is heated until the
piston is twice as high as it was initially. If the atmospheric pressure is 101 kPa,
and the process is 92 percent efficient, how much work is done by the air, and
how much work is required because of dissipation?
Solution.
102 3 Work and Heat
In doing problems that require the work done by the system, we list the
appropriate formula under the sketch, as I have done here. We then need to
evaluate W from the information given. Here, we use Eqs. (3.20) and (3.23)
in order to obtain WS̃R [with WO = −(L/AP )(V2 − V1 )], but we first need to
know the volumes in the initial and final states. These follow from the ideal
gas equation of state
v1 = (0.287 kN m/kg K)(298.15 K)/400 kN/m2 = 0.2139 m3 /kg
Using Eq. (1.23) then gives the initial and final volumes as 0.6418 m3 and
1.2835 m3 , respectively. Then static equilibrium in state 1, p1 = pa + (mP g +
L)/AP , together with Eqs. (3.20), (3.23), and WO give
⎧
⎨ −[(mP g + L)/AP + pa ](V2 − V1 ) = −p1 (V2 − V1 )
WS̃R = −(400 kN/m2 )(1.2835 − 0.6418)m3
⎩
−256.7 kJ
The work done by the system is therefore −(−256.7 kJ)/0.92 = 279 kJ, and
the work dissipated by friction, −WS̃I , is the algebraic sum [from Eq. (3.19)],
22.3 kJ. Note that the piston area could be calculated from the equilibrium
equation.
The mixer/expander shown in Fig. 3.4 is operated by sliding the partition aside
very slowly, and for a very short time; so slowly, and for such a short time, that ẆS̃
is virtually zero. Therefore
WS = −WS̃ ∼ 0 (3.27)
a result we will find useful in problem-solving.
Rotational Work
In this type of motion, the velocity of moving points on the surface of S are given in
terms of the angular velocity ωk, and the distance from the axis of rotation r (refer
to Fig. 3.7) by the formula
vj = ωk × rj
Then Eq. (3.9) can be written as

⎛ ⎞
N
N
ẆS = Fj · (ωk × rj ) = ⎝ rj × Fj ⎠ · ωk
j=1 j=1
3.2 Mechanics 103
Fig. 3.7 A stirring process

is an example of rigid
rotational motion. The
systems shown here are the
stirred fluid S, the propeller
P, the air A, the lubricant L,
and the motor M
where the second equality follows from a property of the mixed triple product.
Accordingly, if we denote the component of the resultant moment along the axis of
rotation by Mk , also called the torque, this is
ẆS = Mk ω (3.28)
In this case we can again use Eq. (3.19) to obtain WS , and here, from Fig. 3.7 we
have
WS̃ = WL + WA + WP + WM
The work done by the propeller, WP , can be obtained from the rotational form of
Eq. (1.34), which you learned in Dynamics; it is
WP = IP (ω22 − ω12 )/2
where IP is the moment of inertia of the propeller for the axis of rotation

IP = ρr 2 d V
V
Therefore if the end states are equilibrium states, or if the angular velocity is the same
in both end states, this term is zero. Moreover, as I described before, the work done
by the lubricant in the bearings and the work done by the air are both irreversible.
Then we have
WS̃R = WM WS̃I = WL + WA
so that in terms of the mechanical efficiency, Eq. (3.24), we get here an application
of Eq. (3.25)
104 3 Work and Heat
WS = −ηM WM WS̃I = (ηM − 1)WM
A typical problem is illustrated in the following example.
Example 3.4 A 1/4 hp motor is used to mix a mass M of a substance in a rigid

tank. At the start of the process, the substance is at pressure p and temperature
T , and the stirring continues for ten minutes, at which point the temperature
is 10% higher. If the efficiency of the process is 87%, how much work is done
on the substance?
Solution.
State 1 M State 2
M1 = M T1 = T T2 = 1.1T1 M2 = M1
(stated) (closed)
p1 = p V2 = V1
(rigid)
The work done by the motor is obtained using Eq. (3.3), WM = {ẆM }
(t2 − t1 ),
WM = (.25 hp)(550 ft lbf/hp s)(600 s) = 82500 ft lbf
The amount of work done on the substance then follows from the equation
listed above; it is 87 percent of WM (71775 ft lbf). As usual, this calculation
requires no state information about the substance. Note that the work done on
the system is the negative of the work done by the system.
3.2.5 Internal Determination of Work Done
The piston moving in a cylinder shown in Fig. 3.4 is special in that there is only
one moving surface whose normal is the same as the direction of motion. Therefore
Eq. (3.16) can be written for this device as
ẆS = −σn AP n · v
where σn is the normal stress from Eq. (1.36). On adding and subtracting pd V /dt
and noting that in this special case Eq. (3.21) is AP n · v = d V /dt, we get
dV dV
ẆS = p − (σn + p)
dt dt
3.2 Mechanics 105
Here the first term on the right hand side is internally reversible [recall the argument
following Eq. (3.21)], whereas the second, for fluid systems, arises as a result of
internal friction, and is dissipative. Thus on writing as usual
ẆS = ẆSR + ẆSI
the reversible power developed is
dV
ẆSR = p (3.29)
dt
This reversible power developed is of importance because it is the maximum a system
can produce
ẆS ≤ ẆSR (3.30)
This follows from the fact that the power dissipated is always positive (or zero
in a frictionless system). The actual power developed is well approximated by the
reversible power when the motion is slow, because the effect of friction decreases with
speed. In addition we sometimes consider frictionless systems in order to estimate
the maximum power that can be developed.
We can progress beyond Eq. (3.29) by considering processes of a very special
type.
Definition 3.1 (Equilibrium Process) An equilibrium process of a thermodynamic

system is a continuous succession of macroscopic equilibrium states.
In our study of thermodynamics we will restrict attention to simple compressible

substances for which an equilibrium process is represented by a space curve
v = vp (t) T = Tp (t) p = pp (t)
which lies in the state surface, pp (t) ≡ p[vp (t), Tp (t)]. In this case V = M vp , so for
a closed system Eq. (3.29) becomes
d vp d vp
ẆS = Mpp = Mp[vp (t), Tp (t)] (3.31)
dt dt
Here we have given the reversible power a special symbol and a name, the equi-
librium power developed, because it plays a special role in thermodynamics. An
equilibrium process is not a real process. According to its definition it is a succession
of equilibrium states, yet equilibrium is defined by an absence of motion. It is thus
an unattainable ideal process, and consequently we need the following axiom, which
is a generalization of our collective experience.
Axiom 3.1 (Equilibrium Power) In an equilibrium process of a thermodynamic
fluid system, the power developed, ẆS , is equal to the equilibrium power, ẆS .
106 3 Work and Heat
Since an equilibrium process is a fiction, no real motion actually develops the equilib-
rium power, however, in some real motions, we can identify an approximate process
curve which lies in the state surface, and the power developed approximates ẆS ;
we call such a process, which is a succession of quasi-equilibrium states, a quasi-
equilibrium process. An important example of such a process occurs when a piston
moves slowly2 in its cylinder, then
ẆS ∼ ẆS (3.32)
Thus in a quasi-equilibrium process, the work done by the system is approximated

by (henceforth we drop the use of the system subscript)
t2
d vp
W=M p[vp (t), Tp (t)] dt
t1 dt
or alternatively using v as parameter, and calling the projection of the curve p in the
T , v plane C [recall Eq. (2.21)]
v2
W=M p[v, TC (v)] d v
v1
The quasi-equilibrium process curve is usually given by using specific volume as

the parameter, and specifying the pressure p = pC (v). In this case
v2
W=M pC (v) d v (3.33)
v1
while the temperature variation is determined implicitly by equating the process

curve equation and the equation of state
pC (v) = p(v, T ) =⇒ T = TC (v)
The work done by a system given by Eq. (3.33) has a simple geometrical interpre-
tation on a p, v diagram. As shown in Fig. 3.8 it is the area under the process curve
multiplied by the system mass. This makes transparent the fact that the work done
in an equilibrium change of state is process dependent, and reversible. Moreover the
work done in an equilibrium cyclic process, one that begins and ends in the same
point, is the area enclosed by the process curve; the system forces are clearly not
conservative.
2 Here slowly means by comparison with the characteristic speeds of the fluid. These depend on the
equations of state of the fluid, but for gases there is only one, the speed of sound, c2 ∝ pv = RT .
Pistons move slowly compared with this speed, about 1100 ft/s in air at 70 ◦ F, so this concept is
useful in practice.
3.2 Mechanics 107
Fig. 3.8 The area under the quasi-equilibrium curve pC (v) is the work done. In the middle
sketch W (t2 , t1 , C− ) = −W (t1 , t2 , C− ), as you learned in Calculus. In the rightmost sketch,
Wc = W (t1 , t2 , C+ ) + W (t2 , t1 , C− ), is the network done in the cycle
Example 3.5 If the process curve for a cyclic process is given implicitly by
the equation
(p − p0 )2 /(p)2 + (v − v0 )2 /(v)2 = 1
where p0 = 30 psi, p = 10 psi, v0 = 0.5 ft3 /lbm, and v = 0.1 ft3 /lbm, how
much work is done per cycle per lbm of matter?
Solution.
This curve is an ellipse (see example 2.1), so the work done per cycle is just
the area enclosed by an ellipse. From elementary geometry this is
W/M = π(p)(v)
The value of W/M is then
W/M = π(10 psi)(0.1 ft3 /lbm)(144 in2 /ft 2 ) = 452.4 ft lbf/lbm
Moreover the work done is independent of p0 and v0 .
We can also prove analytically that the work done is dependent on the process; it
is an interaction, see Sect. 2.3.2. Writing Eq. (3.31) in the full form [see Eq. (2.25)]
dTp d vp
Ẇ = 0 + Mp(vp , Tp )
dt dt
and observing that [see Eq. (2.28)]
∂v 0 = 0 = ∂T p = α/β
the result follows according to the description in Sect. 2.3.2.

108 3 Work and Heat
Once the quasi-equilibrium process curve is known, p = pC (v), the work done by
a system can be found either by evaluating the integral in Eq. (3.33), or by determining
the area under the curve geometrically. In some problems, notably heating or cooling a
gas, this curve is obtained by considering force equilibrium of the piston as described
in Eq. (3.17), along with the assumptions that the velocity is very low and constant.
This means neglecting the dynamic force (in any event the work done by this force
is zero in a change of state process) and the piston friction forces, so Eq. (3.17)
becomes
p = −FO · i/AP + pa + mP gk · i/AP
where k is a unit vector in the local vertical direction. Thus if the external force is a
constant load, FO = −L i, and the motion is upward (k = i) we have
p = pa + (mP g + L)/AP
which is a constant pressure, isobaric, process, p = p1 = p2 = C. The work done in

such a process is simply
W = Mp(v2 − v1 ) (3.34)
If instead, the force is produced by a spring, FO = k( − 0 )i, and the motion is
horizontal, the pressure is
p = pa + kM (v − v0 )/A2P
where v appears as a result of using the relation between and x. This is a linear
process (straight line in the p, v plane), and can be written in terms of the equilibrium
pressures and specific volumes, p = p1 + (p2 − p1 )(v − v1 )/(v2 − v1 ). The work
done is easily obtained from geometry as

p1 + p2
W=M (v2 − v1 ) (3.35)
2
Calculating the work done in this way gives results that are identical with those
described previously (although with 100% efficiency, see Example 3.3).
Example 3.6 One and one third kg of water at 150 ◦ C and 36% quality is
heated in a 30 cm diameter vertical piston cylinder arrangement. At a certain
point the piston encounters stops, however, the heating continues until the
liquid is completely vaporized at 185 ◦ C. If this is a quasi-equilibrium process
in which friction can be neglected, how much work is done by the substance?
3.2 Mechanics 109
Solution.
State 1 State 2
M1 = Tc1 = 150◦ C x2 = 1 M2 = M1
1.33 kg x1 = 0.36 (completely (fixed mass)
vaporized)
Tc2 = 185◦ C
W = M1 p1 (v2 − v1 ) (isobaric until

stops)
In this quasi-equilibrium process with no friction force acting on the piston,

until it reaches the stops the piston is in static equilibrium (it rises at constant
speed) under the same force system that acts on it in State 1. Therefore till this
point the pressure in the water remains the same, and the curve is given by
p = p1 v < v2
This is sketched in the accompanying p, v diagram where the work done is the
area under the curve. Since after the stops are reached the process is isochoric
we find that the work done is that of an isobaric process at the initial pressure.
Now p1 is the saturation pressure at 150 ◦ C which is given in the tables as

0.4758 MPa. The specific volume is calculated from the equation of state of
a liquid–vapor equilibrium mixture, Eq. (2.47) using the tabular values of vf
and vg at 150 ◦ C. This is

[0.001091 + 0.36(0.3928 − 0.001091)]m3 /kg
v1 =
0.1421m3 /kg
The final specific volume v2 is just the saturated vapor value at the final tem-
perature. This is determined by interpolation between the tabular temperature
values of 150 and 200 as
110 3 Work and Heat

0.3928 + [(0.12736 − 0.3928)/(200 − 150)](185 − 150)
v2 =
0.2070m3 /kg
Compiling all these results
W = (1.33 kg)(475.8kN/m2 )(0.207 − 0.1421)m3 /kg
gives the work done as 41.1 kJ.
If we could determine pC (v) by means of arguments about heating, there would

be no need to make assumptions about the motion, as we did previously. Indeed in
that case, we could integrate Newton’s law to determine the motion. This is exactly
what happens (as we shall see later on) with a special class of polytropic processes,
which are defined in general by
pv n = p1 v1n = p2 v2n = C (3.36)
The work done in such a process is most easily found by integration of Eq. (3.33)
v2
W = MC v −n d v
v1
For n = 1 this is
v21−n − v11−n p2 v2 − p1 v1
W = MC =M n = 1 (3.37)
1−n 1−n
which, for an ideal gas, pv = RT , can also be written as
T2 − T1
W = MR n = 1 (3.38)
1−n
For n = 1 the integration produces instead
W = Mpv ln(v2 /v1 ) (3.39)
Equation (3.36) combined with pv = RT shows that this case, n = 1, corresponds to

an isothermal process for an ideal gas.
Example 3.7 An 18 ft3 vessel contains 3 lbm of Methane. The gas is com-
pressed by a piston until the volume is half the initial volume, and the tem-
perature is twice the initial temperature. If the pressure has increased by
70 psi, and the atmospheric pressure is 14.7 psi, what did the pressure gauge
3.3 Thermal Science 111
read prior to the compression? What is the final temperature? If the compres-
sion is polytropic, how much work is done on the gas?
Solution.
State 1 State 2
M1 = p2 = p1 +70 psi M2 = M1
3 lbm (stated) (closed)
V1 = T2 = 2T1 V2 = V1 /2
18 ft3 (stated) (stated)
W = M (p2 v2 − p1 v1 )/(1 − n) (polytropic

process)
This problem was considered in Example 2.9. Therefore here the initial
pressure is 70/3 psi. Consequently the pressure ratio p2 /p1 is 4, and since the
volume ratio v1 /v2 is 2, the polytropic exponent has the value from Eq. (3.36)
n = ln 4/ ln 2 = 2
The work done by the gas is then
W = (280/3 · 9 − 70/3 · 18) lbf ft3 /in2 · 144 in2 /ft 2 /(1 − 2) = −60480 ft lbf
The work done on the gas, −W , is therefore 60480 ft lbf.
3.3 Thermal Science
As noted in Sect. 1.3.9 in connection with temperature measurement, we distinguish

two concepts; the notion of the degree of hotness of a system (its temperature), and
the notion of heating, which occurs when two systems (or two parts of a system)
in thermal contact have different temperatures. Actually there are two aspects of
heating which are of importance. These are the rate of heat transfer to a body, which
measures the amount of heating in terms of the temperature difference between the
two systems, and the rate of heat absorption, which measures the amount of heating
in terms of changes in the temperature (and other properties as well) of either one
of the systems. These temperature changes are known to us from experience to be
such that the heating causes an increase in the temperature of the colder system and
a decrease in the temperature of the hotter one.
112 3 Work and Heat
3.3.1 Heat Transfer
For a process, C, involving heating between two systems, S and S , that takes place
from time t1 to a later time t2 , we say that there has been heat transfer to S in the
amount t2
QS (t1 , t2 , C) = Q̇S dt (3.40)
t1
The value of QS depends on the process, C; like the work done it is an interaction. The
heat transfer to S, QS , is a dimensional quantity with dimension [Q], while the rate
of heat transfer to S, also called the heating of S and denoted by Q̇S , has dimension
[Q/t].
The rate of heat transfer to a system is described locally by the heat flux vector,
q̇(x, t). In terms of this quantity, the rate of heat transfer to S over a small portion,
Aj , of its surface3 is
−nSj · q̇j Aj
where, as usual, nSj is the locally outward pointing unit vector from S. The total rate
of heat transfer to the system, is then just the surface integral

Q̇S = −nS · q̇ dA (3.41)
AS
The heating of S is positive, negative, or zero, depending on the direction of q̇, and
we say in the first two instances that there is heating of or heating by the system,
respectively. In the third case, when there is no heating, we say that the system is
adiabatic, or thermally isolated. This happens, presuming that there is a temperature
difference, when the boundary is insulated.
Conservation of Heating
Imagine a system, S, in contact with a number, N of other systems, Sj , as I indicated

previously in Fig. 3.2. The surface, AS , of S can be divided into portions, which we
denote by AS/Sj , that separate S from each of the surrounding systems Sj , and since
integration is an additive operation, the heating of S is just the sum
Q̇S = Q̇S/S1 + Q̇S/S2 + · · · + Q̇S/SN (3.42)
where the rate of heat transfer to S from Sj is, according to Eq. (3.41)
3 Inaddition to the surface heating which we describe here, there may also be volume heating. We
will not consider this type of heating in this course, but you may very well encounter it in heat
transfer.

Q̇S/Sj = −nS · q̇ dA
AS/Sj
Since the rate of heat transfer to Sj from S is analogously

Q̇Sj /S = −nSj · q̇ dA
AS/Sj
and since for the common surface between S and Sj
nS + nSj = 0
we find that
Q̇S/Sj + Q̇Sj /S = 0 (3.43)
This is the law of conservation of heating, which is the thermal equivalent of the
conservation of power, Eq. (3.12), discussed earlier. Integrating over time, we obtain
a similar result for the heat transfer itself
QS/Sj + QSj /S = 0 (3.44)
For a system U , composed of S and all the Sj as shown in Fig. 3.2, that is ther-
mally isolated, meaning that there is zero heat flux across its boundary, we see from
Eq. (3.41) that Q̇U = 0 (since q̇ = 0 on AU ). Moreover Eqs. (3.42) and (3.43) can
be used to show that
Q̇U = Q̇S + Q̇S1 + · · · + Q̇SN = 0 (3.45)
or in integrated form for the heat transfer
QU = QS + QS1 + · · · + QSN = 0 (3.46)
Modes of Heat Transfer
Heat transfer occurs in several ways, or modes which correspond to different spec-
ifications of q̇. An important mode, conduction, occurs between two solid bodies
directly in contact (see Fig. 3.9), or equivalently, any two parts of a single body. Here
there is no temperature difference at the surface of contact because the temperature
is continuous there, however, there is a temperature gradient (if there was not, all
parts of the body would be at the same temperature, the body would be in thermal
equilibrium, and there would be no heat transfer). Therefore in this case the heat flux
is given by
∂T
q̇ = −k nS
∂n
114 3 Work and Heat
Fig. 3.9 An illustration of the heat transfer problem discussed in the text. The upper temperature
distribution represents convection and occurs when S is filled with fluid, while the lower represents
conduction and occurs when S is a solid
where k > 0 is a nonequilibrium material property, the thermal conductivity, and nS

is the outward pointing normal to S. In one dimension, and constant heat flux over
the heat transfer surface integration of Eq. (3.41) produces
dT
Q̇S/S = kA = −Q̇S /S
dx
with A the surface area. This expression for the heating is called Fourier’s law in
honor of Joseph Fourier who inferred it from experiments, and used it to solve a
variety of conduction heat transfer problems.
Another mode of heat transfer occurs when a body, S, with surface temperature
TS , is heated (or cooled) by a surrounding fluid, S whose temperature far from the
boundary is TS . We use for q̇ at the S/S boundary
q̇ = −hS /S (TS − TS ) nS
where hS /S is a positive constant of proportionality, the heat transfer coefficient. This

expression for the heat flux is known in heat transfer as Newton’s law of cooling.
Integrating this expression for the heat flux, assuming all quantities are uniform over
the heat transfer surface, Eq. (3.41) produces
Q̇S/S = hS /S A(TS − TS ) = −Q̇S /S (3.47)
Since hS /S > 0 this equation states that the heat transfer to S is positive when its
temperature is less than that of S in accordance with our experience. Equation (3.47)
is used to model the heat transfer process which occurs when heat is transferred via
the motion of a fluid. Common examples of this phenomenon, called convection, are
the heating of room air by a radiator, and the chilling effect of wind in winter.
Applying the test of Eq. (3.7) equation shows that both these modes of heat transfer
are irreversible (they are independent of velocity), however, they are not dissipative
since the heating can be positive or negative. Experience shows that all modes of
heat transfer are irreversible.
We can use these ideas to study the simple heat transfer problem shown on the left
in Fig. 3.9. There, systems H , L, and S form a thermally isolated combination in which
heat is transferred from H at temperature TH , to L at temperature TL < TH through
S. If we regard S as a fluid system with temperature TS far from its boundaries, we
have from Eq. (3.47)
−Q̇H = Q̇S/H = hS/H A(TH − TS ) Q̇L = Q̇L/S = hS/L A(TS − TL )
We imagine systems H and L to be very large so that their temperatures do not

change significantly as a result of this heat transfer process. On the other hand TS
varies until it reaches a value for which there is no longer any heating of S, Q̇S = 0,
at which point all changes cease. In this case, which we call steady heat transfer,
Eq. (3.45) shows that effectively H is heating L, or Q̇L = −Q̇H . Using the equations
for −Q̇H and Q̇L then allows us to solve explicitly for TS . This is most easily done by
writing (TH − TL ) − (TS − TL ) for TH − TS in −Q̇H above and solving for TS − TL .
We get
TS − TL = [hS/H /(hS/L + hS/H )](TH − TL )
and therefore the rate of heat transfer, independent of TS , is
Q̇L = ht A(TH − TL ) = −Q̇H (3.48)
where 1/ht = 1/hS/L + 1/hS/H .

When S is a solid body undergoing steady conduction the temperature is a linear
function of position (because Q̇S/S is the same at every location) so
Q̇L = (k/D)A(TH − TL ) = −Q̇H
This is Eq. (3.48) with ht = k/D.

When a solid wall of thickness D and thermal conductivity k separates a hot fluid,
whose temperature is TH far from the wall, from a cold fluid, whose temperature is
TL far from the wall, Eq. (3.48) describes this heating process with
1/ht = 1/hS/L + 1/(k/D) + 1/hS/H
This discussion indicates that the quantity 1/ht A physically describes a total resis-
tance to heat transfer, where every element adds its own resistance to the sum.
Equations (3.47) and (3.48) are rarely used in thermodynamics; they form the
basis for the subject of heat transfer. Although I have discussed them here mainly to
help you understand something about heating, we will use them from time to time
later in this course.
116 3 Work and Heat
3.3.2 Heat Absorption
When we heat a system, we expect to observe changes in all its properties; a fact
that is the basis for the temperature measurement schemes that were discussed in
Sect. 1.3.9. Accordingly we introduce the following definition.
Definition 3.2 (Heat Absorption) In a closed thermodynamic system, S, with D
independent variables, the rate at which heat is absorbed by the system, Q̇S , in an
equilibrium process, yj = yjp (t) with 1 ≤ j ≤ D, is a dimensional quantity, [Q/t],
which consists of a linear combination of the rate of increase of each of the indepen-
dent variables multiplied by a coefficient which is an extensive property of state

D
dyjp
Q̇S = Yj [y1p (t), . . . , yDp (t), M ]
j=1
dt
Before investigating the relationship between the twin concepts of heat transfer and
heat absorption, we first need to discuss some characteristics of heat absorption as
we did previously for heat transfer.
Our principal interest is in simple compressible substances, for which equilibrium
is specified by two intensive variables. Taking these as T and v, the definition gives
(here and in the future we drop the system subscript as long as the context prevents
confusion regarding the system under consideration)
dTp d vp
Q̇ = Mcv + M λv (3.49)
dt dt
Here there are two new equilibrium properties, cv (v, T ), the specific heat at con-
stant volume, and, λv (v, T ), the latent heat at constant volume. We can see from
this equation that the dimensions of specific heat are [Q]/[M T], and for latent heat
[Q]/[L3 ]. The first term of Eq. (3.49) describes a part of the heating that raises the
temperature of the system; it is called the sensible heating. The second term, which
goes to increase the volume of the system is called the latent heating. Under normal
circumstances, both cv and λv are positive; namely, when a system is heated at con-
stant volume its temperature increases, and when it is heated at constant temperature,
its volume increases (for an exception see the footnote in Sect. 1.3.10).
There is nothing special about the particular choice of independent variables made
above. We could have used T and p, in which case we would have obtained
dTp dpp
Q̇ = Mcp + M λp (3.50)
dt dt
where cp is the specific heat at constant pressure, and λp is the latent heat at constant
pressure.
The four properties defined in Eqs. (3.49) and (3.50) are not independent, for on
using the derivative of the equation of state, [see Eq. (2.18) and Sect. 2.4.1]
d vp dTp dpp
= vp α − vp β (3.51)
dt dt dt
in Eq. (3.49) we obtain
dTp dpp
Q̇ = M (cv + vp αλv ) − M vp βλv
dt dt
Equating this with Eq. (3.50), and noting that the temperature and pressure derivatives
can be varied independently of one another, we obtain
cp = cv + vαλv (3.52)
λp = −vβλv (3.53)
From these we deduce that normally cp is greater than cv , and λp < 0. The latter
implies that heating at constant temperature is accompanied by a decrease in pressure.
It is usual to take the independent properties to be the two specific heats, Eqs. (3.52)
and (3.53) then can be solved for the two latent heats
λv = (cp − cv )/vα (3.54)

λp = −(cp − cv )β/α (3.55)
An exception to this rule occurs for states that are combinations of phases since then
p and T are not independent variables, and therefore cp and λp have no meaning
[according to Definition 3.2 Eq. (3.50) does not exist in this situation]. This solution
also makes no sense when α = 0, but we will take no interest in these isolated points
(see footnote in Sect. 1.3.10) in this course.
The heat absorbed is related to the rate of heat absorption in the usual way
t2
Q= Q̇ dt (3.56)
t1
and is found by substituting Eq. (3.49) into (3.56) and integrating

t2
dTp d vp
Q=M cv [vp (t), Tp (t)] λv [vp (t), Tp (t)] dt
t1 dt dt
This can be written alternatively by parameterizing with respect to T and v, and

calling the projection curve C [refer to Eq. (2.21)], as we did with work
T2 v2
Q=M cv [vC (T ), T ] dT + M λv [v, TC (v)] d v (3.57)
T1 v1
118 3 Work and Heat
Although it is not as obvious as it was with the work done in a quasi-equilibrium

process, Eq. (3.33) (see also Sect. 3.2.5), the value given by Eq. (3.57) is also process
dependent. Experimental determinations of the specific heats reveal that
∂v cv = ∂T λv = ∂T [(cp − cv )/vα] (3.58)
so that the heat absorbed is, like the work done, an interaction,4 and is not expressible
as a difference of property values (see Sect. 2.3.2).
3.3.3 Measurement of Specific Heats
The existence of processes for which the heating is independent of the mode of heat
transfer allows us to define a scale for the measurement of [Q] in terms of equilibrium
properties. Thus in a constant pressure, equilibrium process the heat absorbed can
be written as a single integral, after using Eq. (3.50) in Eq. (3.56)
T2
Q=M cp (p, T ) dT
T1
If cp does not vary significantly over the temperature range of the process, we can
regard it as a constant, and the integral is simply
Q = Mcp (T2 − T1 ), p = p1 = p2 = C (3.59)
This equation in conjunction with a standard calorimetric substance, defines a

unit for the measurement of heat. Thus a unit quantity of heat absorbed is the amount
required to raise the temperature of a unit mass of the standard substance by one
degree in a constant pressure process. The standard substance is commonly taken to
be water, and two scales are used; the British Thermal Unit (Btu), which is the amount
of heat absorption required to raise 1 lbm water from 58.5 to 59.5 degrees Fahrenheit
at 1 atmosphere, and the kilocalorie (kcal), which is the amount of heat absorption
required to raise 1 kg water from 14.5 to 15.5 degrees Celsius at 1 atmosphere. The
implication of these definitions is that, for water at the mean temperatures,
cp = 1 kcal/kg ◦ C cp = 1 Btu/lbm ◦ F
The conversion between these two units is accomplished by equating the two- dimen-
sional values of cp
4 Due to difficulties in determining specific heats accurately, early investigators of heating were led
into error at this point. Accordingly they believed that in an arbitrary change of state a determined
amount of heat was absorbed, equal to the change in a system property they called the amount of
caloric This error was corrected long ago, however, vestiges remain in our use of the word heat as
a noun, as though heat was a state function like pressure, temperature, and specific volume.
1 kcal/kg ◦ C = 1 Btu/lbm ◦ F × 1 lbm/0.45359237 kg × 1.8 ◦ F/ ◦ C

1 kcal = 3.9683 Btu
Measurement of the constant pressure specific heat of other substances, and of

water at other temperatures and pressures, is done through a calorimetric process,
in which a hot body, H at temperature TH , and a cold body, C at temperature TC
are brought into diathermal contact. They are thermally isolated from other bodies,
and they reach a common final temperature T through a constant pressure quasi-
equilibrium process. In order to justify this procedure we make use of an axiom.
Axiom 3.2 (Calorimetry) In an equilibrium process of a thermodynamic system

the rate of heat transfer to the system, Q̇, is equal to the rate of heat absorption by
the system, Q̇.
As mentioned previously, an equilibrium process is a fiction, indeed the words equi-
librium and process have opposite meanings. However, Axioms 3.1 and 3.2 are useful
because we can infer from them that in a quasi-equilibrium process, Ẇ ∼ Ẇ, and
Q̇ ∼ Q̇. Moreover, the closer the real process comes to the ideal, the more accurate
are the approximations of Ẇ and Q̇.
The calorimetric process is a four-state process that describes an interaction of
two systems. We can cast the information that is provided in our usual form
where the rate of heat absorption has been used as an approximation for the heating of
each system. Writing the Q expressions in the adiabatic (thermally isolated) condition
in terms of the common final temperature, T
120 3 Work and Heat
MH cpH (T − TH ) + ML cpL (T − TL ) = 0
If the masses and all temperatures are known, this can be easily solved for the specific
heat of one of the bodies in terms of the other. In this way cp for water in the liquid state
has been determined to be nearly one over a wide range of pressure and temperature.
Moreover values of cp for other substances have also been measured. For liquids,
solids, and ideal gases the cp values are found to depend only on temperature. Some
of these values are listed in the tables provided.
Alternatively, if the masses, specific heats, and initial temperatures are known,
we can easily determine the final, equilibrium temperature [recall the substitution in
obtaining Eq. (3.48)]
T − TL = (TH − TL )/(1 + r) (3.60)
where r = ML cpL /MH cpH is called the thermal mass ratio.
Example 3.8 A 50 lbm block of copper at 400 ◦ F is dropped into a bucket of

water whose volume is 1 ft3 , and is at 60 ◦ F. What is the final temperature?
Solution.
The thermal mass ratio for this example is
(1 ft3 )(62.16 lbm/ft3 )(1 Btu/lbm R)

r= = 13.51
(50 lbm)(.092 Btu/lbm R)
so that Eq. (3.60) gives for the Fahrenheit temperature
Tf = 60 ◦ F + (400 − 60) ◦ F/(1 + 13.51) = 83.4 ◦ F
Notice that here we can calculate the Fahrenheit temperature directly because
the equation only involves temperature differences [see Eq. (1.49)].
A plot of Eq. (3.60) is shown in Fig. 3.10. Two important limits are clearly seen
in this figure.
lim T = TH lim T = TL
r→0 r→∞
Each of these limits relays the same physical content, namely that whenever the
thermal mass of one system is very much larger than the other, the final temperature
is very close to the initial temperature of the system with the larger thermal mass.
Such a system, called a thermal reservoir, can transfer heat without significant change
in its temperature, in other words, isothermally. Conversely, a thermometer should
have a thermal mass much smaller than that of the system whose temperature is to
be measured, so that the indicated temperature is essentially that of the undisturbed
system.
Fig. 3.10 Dimensionless

representation of Eq. (3.60)
Measurement of the constant volume specific heat is similarly done by means of

a constant volume, quasi-equilibrium process for which Eq. (3.57) is just
T2
Q=M cv (v, T ) dT
T1
For a small temperature range over which cv can be regarded as constant, the integral
can be simply evaluated as
Q = Mcv (T2 − T1 ), v = v1 = v2 = C (3.61)
In a calorimetric process in which a mass Mv of one substance undergoing a constant

volume process from an initial temperature T1v , is placed in thermal contact with a
mass Mp of another substance undergoing a constant pressure process from an initial
temperature T1p , cv of the first substance is related to cp of the second by another
four-state process. If the final temperature is T the result is
Mv cv (T − T1v ) + Mp cp (T − T1p ) = 0
or solving for cv
cv = −Mp (T − T1p )/[Mv (T − T1v )]cp
For liquids, solids, and ideal gases the cv values are also found to depend only on
temperature. They are listed along with cp for some selected substances in your
tables.
In similar fashion calorimetric processes can be used to measure other thermal
quantities of interest. For example, measurements have shown that 1 kcal melts 12.55
grams of ice at 1 atmosphere pressure (see Sect. 1.3.9).
122 3 Work and Heat
3.3.4 External Determination of Heat Transfer
In a general process, the heat transfer to a system, S, can be determined by surrounding

the system with another system, M , for which the heat transfer can be approximated
by the heat absorption, and hence measurable by the methods we have just discussed.
Accordingly, if we denote the heat transfer to the system by the symbol QS , and the
heat absorbed by the measurement system with QM , conservation of heating for this
thermally isolated combination, in the form of Eq. (3.46), provides the relation (M
undergoes a constant pressure process)
QS = −QM = −MM cpM (T2M − T1M ) (3.62)
Example 3.9 A 1/4 hp motor is used to mix a mass M of a substance in a rigid

tank. At the start of the process the substance is at pressure p and temperature
T , and the stirring continues for ten minutes, at which point the temperature
is 10% higher. If the tank is in thermal contact with a bath containing 3 lbm
water whose temperature increases by 20 ◦ F during the process, and there is
no other heat transfer from the tank, what is the amount of heat transfer from
the substance?
Solution.
The amount of heat transfer to the substance is given by Eq. (3.62) which
is listed above. Since the measurement substance is water, we get
Q = −(3 lbm)(1 Btu/lbm R)(20 R) = −60 Btu
This is the heat transfer to the substance, therefore the heat transfer from
the substance is 60 Btu. Notice that this calculation was done using no state
information about the substance.
3.3.5 Internal Determination of Heat Transfer
It is not possible to evaluate Eq. (3.56) in simple terms for processes, other than the
constant pressure and constant volume ones we have done already, without making
use of the equation of state, and other simplifying assumptions about the behavior
of the specific heats. However, you can obtain some insight into the relationship
between heat and work from such an exercise. In addition, I will use the discussion
to describe several processes that are important in applications.
• In an isothermal process,
p = pT (v) T = T1 = T2 = C
the heat absorbed is given by Eq. (3.56), using Eq. (3.49), with Eq. (3.54)
v2
cp − cv
Q=M dv (3.63)
v1 αv
If the specific heats and coefficient of thermal expansion do not vary significantly
over the volume range of the process (this is true to a high degree of accuracy for
liquids, solids, and ideal gases), we can regard them as constant in Eq. (3.63) and
obtain
cp − cv v2
Q=M ln , T = T1 = T2 = C (3.64)
α v1
• An adiabatic process, Q̇ = 0, is the thermal equivalent of a zero power process,

Ẇ = 0. Moreover an equilibrium adiabatic process, Q̇ = 0 is the thermal equiv-
alent of an isochoric process, v = C, for which Ẇ = 0. If we parameterize this
process curve using v, so that
p = ps (v) T = Ts (v)
we obtain from Eqs. (3.49) and (3.50) [with Eqs. (3.54) and (3.55) for the latent
heats]

dTs cp − cv d vp
Q̇/M = cv + =0
dv vα dt

dTs β dps d vp
Q̇/M = cp − (cp − cv ) =0
dv α d v dt
Now Dt vp = 0, so on eliminating the derivative dTs /d v between the two terms in

square brackets, we obtain the slope of the process curve in the p, v plane
dps cp 1
=−
dv cv vβ
124 3 Work and Heat
For an ideal gas, β = 1/p, with constant ratio of specific heats (we call such gases
perfect gases), k = cp /cv , this differential equation can be integrated

1 dps dv
d v = −k
ps d v v
to give the equation of the equilibrium adiabat
ps = C/v k
This is a polytropic process, Eq. (3.36), with exponent n = k [see the comment
just before Eq. (3.36)]. Since the process occurs in an ideal gas, the temperature
variation is
Ts = ps v/R = C/Rv k−1
• In a polytropic process, T = C/Rv n−1 , of an ideal gas, α = 1/T , with constant

specific heats, cp /cv = k, the heat absorption is, on integrating Eq. (3.49) with
Eq. (3.54) v2
Q/M = cv (T2 − T1 ) + [(cp − cv )/R]C v −n d v
v1
Evaluating the integral here [recall Eq. (3.38)] and factoring produces
n−k
Q=M cv (T2 − T1 ) n = 1 (3.65)
n−1
We will return to these matters again in Chap. 4 when we try to find out exactly how
work and heat are related to each other.
3.4 Exercises
In each of the following exercises where appropriate, indicate the change of state (or
process) on a suitable projection plane of the state surface. For substances that can
undergo a phase change sketch the saturation curve as well in this diagram.
Section 3.2.4
3.1 A shock absorber is forced with a constant magnitude force of 100 lbf, and
direction the same as its velocity, at a frequency of 50 Hz. If the motion is linear in
time, reversing at the maximum and minimum displacement, with a stroke of 4 in,
how much work is done by the shock in each cycle of operation? What is the average
power dissipated? (−800 in lbf, 6.06 hp)
3.4 Exercises 125
3.2 A 3/8 hp motor operates a shaker filled with a substance to be mixed. If the
motor runs for 10 minutes, and the mechanical efficiency is 92%. How much work
is done on the substance in the shaker?
3.3 One lbm of saturated liquid R-12 at 20 ◦ F is heated in a 5 in diameter vertical

piston cylinder arrangement. At a certain point the piston encounters stops, however,
the heating continues until the liquid is completely vaporized at 165 ◦ F. If the piston
is in equilibrium at the initial state, the dissipation is 75 ft lbf, and there is no output
device connected to the piston, how much work is done by the substance, and what
is the efficiency? (685 ft lbf, 89%)
3.4 One kg of steam is initially contained in a 0.1 m diameter vertical cylinder at 1

MPa and 300 ◦ C, is compressed to 95% of its original volume by putting an additional
118 kg mass on the piston. If the mechanical efficiency of the process is 85%, how
much work is done on the steam?
3.5 A 1 kW motor is used to mix a substance. The amount of work dissipated is

5000 J and the mechanical efficiency is 80%. How long did the motor run, and how
much work was done on the substance? (25 s, 20 kJ)
3.6 A 1/2 kW motor operates a mixer filled with some substance. If the motor runs
for 12 minutes, and the mechanical efficiency is 90%, how much work is done on
the substance in the mixer?
Section 3.2.5
3.7 Obtain an expression for the work done in the cyclic process composed of an
isochoric process at vL from pH to pL < pH , an isobaric process at pL from vL to
vR > vL , and a linear process back to the first point.
3.8 One lbm of saturated liquid R-12 at 20 ◦ F is heated slowly in a 5 in diameter

vertical piston cylinder arrangement. At a certain point the piston encounters stops,
however, the heating continues until the liquid is completely vaporized at 165 ◦ F.
How much work is done by the substance? (610 ft lbf)
3.9 One lbm of saturated liquid R-12 at 20 ◦ F is heated slowly in a 5 in diameter

vertical piston cylinder arrangement in which a linear spring is compressed by the
piston as the substance expands. The heating continues until the liquid is completely
vaporized at 165 ◦ F. How much work is done by the substance? What is the value of
the spring constant?
3.10 One kg of argon expands in a polytropic process (n = 1) from 2 MPa, 400 ◦ C

to 0.8 MPa. How much work is done by the argon? What is the final temperature?
(128 kJ, 400 ◦ C)
126 3 Work and Heat
3.11 One kg of steam expands in a polytropic process (n = 1) from 2 MPa, 400 ◦ C

to 0.8 MPa. How much work is done by the steam? What is the final temperature?
3.12 One kg of steam expands from 2 MPa, 400 ◦ C to 0.8 MPa. If the process is
isothermal, how much work is done by the steam? Use some numerical integration
scheme in order to obtain your answer. (285 kJ)
3.13 One kg of methane expands in a polytropic process (n = 1.2) from 2 MPa,

400 ◦ C to 0.8 MPa. How much work is done by the methane? What is the final
temperature? If in the initial, constrained equilibrium, state the piston was pinned
in place, and after the pins were removed it moved, under the action of gravity and
atmospheric pressure, to its final equilibrium state, how much work is done by the
surroundings, and what is the efficiency of the process?
3.14 One-quarter kg of air is contained in a cylinder at 100 kPa and 30 ◦ C. The

piston is impacted by a force that compresses the air to 1/10 its original volume in
a change of state that is polytropic with exponent 1.35. How much work is done by
the air (recall Exercise 2.24)?(−308 kJ)
Section 3.3.1
3.15 A hot fluid, TH , heats a cold one, TL through a solid wall of thickness D and
thermal conductivity k. Show that the heat transfer is governed by Eq. (3.48) with
1/ht = 1/(k/D) + 1/hS/L + 1/hS/H , and that the solid wall temperature at the hot
fluid interface is TS/H = TH − (ht /hS/H )(TH − TL ) whereas at the cold fluid interface
it is TS/L = TL + (ht /hS/L )(TH − TL ).
3.16 Use the results of the previous problem to determine the surface temperatures
of an automobile windshield (of thickness 4 mm) when the defroster air temperature
is 40 ◦ C and the exterior temperature is −10 ◦ C. The interior and exterior heat transfer
coefficients are 30 w/m2 K and 65 w/m2 K, respectively, and the thermal conductivity
of glass is 1.4 w/m K. (7.7 ◦ C, 4.9 ◦ C)
3.17 The wind chill is related to increased heat transfer from exposed skin, S, to
the surrounding atmosphere, A, on a windy day. Consider a layer of tissue (k = 0.2
Btu/hr ft R) 1/8 in thick whose inner surface is maintained at 98.6 ◦ F. On a calm day
hS/A = 7.9 Btu/hr ft2 R while on a windy day it is 21 Btu/hr ft2 R. If the ambient
temperature is 25 ◦ F, what is the outer surface skin temperature on the calm and windy
days? What calm day ambient temperature produces the same cooling as 25 ◦ F on
the very windy day?
3.18 The window of a self-cleaning oven is made of plastic with thermal conduc-
tivity 0.025 w/m K. The oven temperature during cleaning is 400 ◦ C, and the kitchen
temperature is 25 ◦ C. If the heat transfer coefficients at both the inside and outside
window surfaces are 25 w/m2 K, what is the minimum window thickness that will,
for safety’s sake, produce an outside plastic surface temperature less than or equal
to 50 ◦ C? (1.3 cm)
3.4 Exercises 127
Section 3.3.2
3.19 Obtain an expression for Q̇ using v and p as intensive variables, by eliminating

dTp /dt between Eqs. (3.49) and (3.50).
3.20 Use Axiom 3.2 along with Eqs. (3.47) and (3.50) to derive the equation that
governs the temperature of a body cooking in an oven. Make the assumption that the
cooking process is a constant pressure process to obtain
dT
t∗ = TH − T
dt
where t ∗ = Mcp /ht A is the characteristic time for cooking, and TH is the constant
oven temperature.
3.21 Integrate the differential equation of the previous problem with the initial
condition that initially, t = 0, the body is at temperature T0 , to obtain the result
∗
T = TH − (TH − T0 )e−t/t
Plot this result in the form (T − TH )/(T0 − TH ) versus t/t ∗ , using any convenient
software, over the range 0 ≤ t/t ∗ ≤ 2.
3.22 If t ∗ in the previous problem is 1 hr, the oven is set at 375 ◦ F, the initial body
temperature is 70 ◦ F and is done when its temperature is 260 ◦ F, what is the cooking
time? If the body is placed in the oven directly from the refrigerator, T0 = 40 ◦ F,
what is the cooking time? (58 min, 64 min)
Section 3.3.3
3.23 Two lbm of a solid material at 100 ◦ F is placed in thermal contact with 0.8 lbm
of water at 40 ◦ F. If the final common temperature is 62 ◦ F, what is cp of the solid?
(0.232 Btu/lbm R)
3.24 A 5 lbm piece of steel at 600 ◦ F is dropped into a bucket of machine oil at
70 ◦ F to quench it. If the final temperature is to be not higher than 110 ◦ F, what is
the minimum size bucket needed in gallons?
Section 3.3.4
3.25 You want to design a vessel to measure 120 Btu resulting from a process. If
you decide to use water in the vessel, and allow for a 20 ◦ F temperature rise, what
is the required volume? What would the volume be if you used ethyl alcohol in the
vessel? (0.0965 ft3 , 0.214 ft3 )
3.26 Two lbm of water in a calorimeter decreases in temperature from 80 ◦ F to

66 ◦ F. How much heat was transferred from the water?
128 3 Work and Heat
Section 3.3.5
3.27 One half lbm of air expands isothermally from 80 psi, 100 ◦ F to 20 psi. How
much work is done by the air, how much heat is absorbed, and what is the ratio of
work done to heat absorbed? (20700 ft lbf, 27.2 Btu, 770 ft lbf/Btu)
3.28 Show by using Eqs. (3.39) and (3.64), that Q = [(cp − cv )/R]W in an isother-
mal change of state of an ideal gas.
Chapter 4
The First Law
4.1 Introduction
In the last chapter, you learned how to calculate the work done by and the heat
transfer to a system that undergoes a specified change of state. On the other hand,
you could not calculate the change of state caused by given values for the work
and heat interactions. Furthermore, although you learned how to calculate the work
done and heat transfer independent of one another, experience tells us that they are
intimately related. Indeed there is a fundamental relation between the work done
by a system, the heat transfer to the system, and its change of state. Exploring and
exploiting this relation is one of the basic concerns of thermodynamics.
Although there seems to be little connection between either the general expres-
sions for power developed and heating, Eqs. (3.10) and (3.41)

Ẇ = −τ n · v d A Q̇ = −n · q̇ d A (4.1)
A A
or the versions specialized for equilibrium processes, as in Eqs. (3.31) and (3.49)
dv dT c p − cv dv
Ẇ = M p Q̇ = Mcv +M
dt dt vα dt
when we specialize Q̇ further, for an ideal gas pv = RT (so that α = 1/T )
dT c p − cv dv dT c p − cv
Q̇ = Mcv +M p = Mcv + Ẇ
dt R dt dt R
For constant specific heats, c p and cv , an additional simplification we call a perfect
gas, rearrangement of this expression produces
dU
K M Q̇ = + Ẇ (4.2)
dt
130 4 The First Law
where K M = R/(c p − cv ), and U = M K M cv T is an extensive variable with the

dimension of energy. Although K M here is a property with dimensions [E]/[Q], J.R.
Mayer, around 1840, hypothesized that it is a constant with the same value for all
perfect gases. Mayer’s hypothesis makes K M ≡ k J , a dimensional constant, (recall
Sect. 1.2.1), and means physically that for ideal gases with constant specific heats,
perfect gases, heat absorption is equivalent to mechanical work.
4.2 Internal Energy and the Energy Equation
Equation (4.2), which applies only to a perfect gas undergoing an idealized process
(equilibrium process), is nevertheless a prototype of the relationship we seek. Indeed
experiments have shown that Eq. (4.2), augmented by Mayer’s hypothesis, and suit-
ably generalized, applies to all thermodynamic systems and processes. Known as the
first law of thermodynamics, it is one of the fundamental principles upon which all
physical theories of matter are based. As usual, we formulate an axiom from which
we can deduce the known phenomena
Axiom 4.1 (First Law) For all closed thermodynamic systems that admit macro-
scopic equilibrium states there exists an extensive property, U, called the internal
energy such that:
1. in an equilibrium state, the system internal energy is, apart from a constant, a
function of state and,
2. during an arbitrary process, the sum of the rate of increase of internal energy
and the total power developed by the system is equal to the product of the total
rate of heat transfer to the system and a dimensional constant
dU
+ Ẇ T = k J Q̇ T (4.3)
dt
For non-rotating mechanical systems Ẇ T = Ẇ + d K c /dt + dφc /dt, where K c and

φc are the kinetic and potential energies of the mass center respectively, Ẇ is given
by Eq. (4.1), and Q̇ T is the sum of the surface heating given in Eq. (4.1) and any
volume heating that may exist (although in this course we will not consider volume
heating). The equation can be applied to other types of systems by incorporating the
appropriate terms (e.g. electrical, magnetic) in the expression for total power.
Credit for proposing this axiom is given to Rudolf Clausius, William Macquorn
Rankine, and William Thomson (Lord Kelvin), who all used some form of it within
a 12 month period (in 1850-51). In addition to Mayer’s hypothesis, it was founded
on the many various mechanical and electrical experiments performed by James
Prescott Joule between 1840 and 1850.
The first part of Axiom 4.1 specifies, U as a state function. This means that for a
single phase, simple compressible substance in a state of macroscopic equilibrium
there is an equation of state of the form
4.2 Internal Energy and the Energy Equation 131
U = U (v, T, M) = Mu(v, T ) (4.4)
where u is the specific internal energy, an intensive variable. The choice of the
intensive independent variables here is arbitrary, and was made for specificity. We
could easily have chosen p in place of either one in view of the mechanical equation
of state, Eq. (2.8).
In order to use Eq. (4.3) in the solution of practical problems, we need to know
the energetic equation of state, u(v, T ), just as we know the mechanical equation of
state from our study in Chap. 2. We will undertake this development later in Sect. 4.3.
Kinetic Model for Internal Energy
We can learn something about internal energy by resorting to our previous model
[recall Eqs. (1.41) and (1.50)] of a gas composed of noninteracting particles. In such a
model, if the particles have no internal structure, the only energy they possess is their
translational kinetic energy, m p c2 /2, so we have for the energy of the N p particles
1
U= N p m p c2 + constant (4.5)
2
What is the meaning of the arbitrary constant associated with U ? Like all other forms
of energy, there is no physical meaning to the internal energy itself which is a relative
quantity (see Sect. 1.2.2). Only differences in energy, which are manifested physically
as work done or heat transfer, are absolute. Moreover, only energy differences appear
in Axiom 4.1, therefore the value of the constant is inconsequential. Recall also
from Mechanics that only differences in kinetic and potential energy have physical
meaning.
4.2.1 The Equivalence of Work and Heat
In Axiom 4.1, k J is a constant with dimensions [E/Q] = [ML2 /t2 Q]. Although the
relationship between heat and work was originally written in this way, when they
were thought to be very different quantities, and they were measured in terms of
very different units, it is nowadays customary to regard heat as a derived dimension
given by, [Q] = [E] = [ML2 /t2 ] (refer to Sect. 1.2.1); in physical terms we think of it,
like work, as simply another form of energy transfer. Accordingly we set k J equal
to unity, and write Eq. (4.3), the energy evolution equation, in the form
dE
= Q̇ − Ẇ (4.6)
dt
where, E = U + K c + φc , is the total system energy. Equation (4.6) is the form in

which the equation is normally encountered in thermodynamics. We can obtain from
it several important general truths.
132 4 The First Law
Theorem 4.1 (Conservation of Energy) During an arbitrary motion of a mechan-

ically and thermally isolated closed thermodynamic system, its total energy is con-
served.
Proof. Integrating Eq. (4.6) from an initial time, t1 to an arbitrary later time, t,
produces
E(t) − E(t1 ) = Q − W
In a mechanically isolated system, W is zero [see Eq. (3.15)], and in a thermally

isolated system Q is zero [Eq. (3.46)]. Using these values, we find that the total
energy is always the same as it was originally.
This result is philosophically as well as practically important, however in order to

apply it we need to know more about the internal energy function. The following
corollaries are more immediately useful.
Corollary 4.1 In any process of a closed, thermally isolated system for which the
total energy is constant, the work done by the system is zero.
A steady-state or continuous process is one for which all system properties are
constant in time. Therefore this corollary applies to continuous processes.
Corollary 4.2 In any process of a closed, thermally isolated system for which the
total energy is periodic, the work done by the system in any period is zero.
A cyclic process is one for which all system properties are periodic in time. Therefore
this corollary applies to cyclic processes.
Ingenious inventors continually propose machines that are intended to produce
limitless amounts of useful work, on a continuous or cyclic basis, with no supply of
or consumption of energy. Such machines are called perpetual motion machines of
the first kind because they supposedly never stop for fuel. No such machine has ever
worked as claimed, a fact that is consistent with the result of these corollaries.
Corollary 4.3 In any process of a closed system for which the total energy is con-
stant, the power developed by the system is equal to the rate of heat transfer to the
system. In any time interval, the work done by the system is equal to the heat transfer
to the system.
The result of this corollary

Ẇ = Q̇ (4.7)
applied to continuous (steady-state) processes is important both for practice and

problem solving.
Corollary 4.4 In any process of a closed system for which the total energy is peri-
odic, the average power developed by the system is equal to the average rate of heat
transfer to the system. In any time interval equal to the period the work done by the
system is equal to the heat transfer to the system.
Fig. 4.1 A schematic representation of Joule’s experiment which is described in the text
The results of this corollary

{Ẇ } = { Q̇} (4.8)
and
W (t1 , t1 + tc , C) = Q(t1 , t1 + tc , C) (4.9)
where t1 is an arbitrary initial time, tc is the cycle time, and C is a cyclic process, are
the cyclic analogue of the previous results for continuous processes.
The first experiments that established numerical values in the conversion relation
between work and heat were done by J. P. Joule; they were instrumental in establishing
Eq. (4.9). He arranged a cyclic process, see Fig. 4.1, in which, at the outset, a falling
weight performed a measured amount of rotational work on a fixed quantity of water.
After noting an increase in temperature, he subsequently arranged for a process of
heat transfer from the water to a calorimeter that was continued until the temperature
of the stirred water was the same as it had been initially. According to Eq. (4.9),
the measured amount of work done by the falling weight was equal to the measured
amount of heat transferred to the calorimeter, and this equality produced the results
(we give here the best modern values)
1Btu = 1.055 kJ = 778.2 ft-lbf 1 kcal = 4.187 kJ
The same conversion relations are obtained in this experiment independent of the
amount of work done and heat transferred, as required by Corollary 4.4. Moreover,
Joule’s experiment has been repeated many times by others, and variants of his
original experiment have been carried out using other sources of energy, all with the
same result.
From the size of the numbers you can appreciate why the development of the
steam engine had such a historic impact on civilization. According to the conversion
relation, cooling one lbm of water by one degree Fahrenheit is equivalent to the
134 4 The First Law
amount of work you would do in lifting about 110 lbm over your head! In the
SI system, the unit used to express all quantities related to heat, is the Joule. The
kilocalorie has almost completely disappeared; one of its remaining uses is in labeling
the energy content of foods, where it is known as a calory (Cal). In the English System,
the Btu is still an acceptable subsidiary unit, but it is now defined in terms of the
Joule by the conversion given in the previous paragraph. Consequently, c p of water
at 59 ◦ F is no longer equal to 1 Btu/lbm ◦ F by definition (Sect. 3.3.3), and the Btu is
no longer a substance-dependent unit.
Equation (4.6) can be written in an alternative form, that has the advantage of
promoting physical insight. In this form, the equation is an energy balance
dE
= Ė T + Ė G (4.10)
dt
which expresses the fact that even when it is changing, energy is neither created
nor destroyed. Rather, the increase of energy of a fixed amount of matter is equal
to the rate at which it is transferred across the surface, Ė T , plus the rate at which
it is generated within the volume, Ė G . Two physical mechanisms contribute to the
rate of energy transfer, the heating, and the power developed. There is no volume
generation term (it is zero) in our applications, however we can conceive of cases
where it is nonzero (for example, see the footnote in Sect. 3.3.1). Such balance, or
conservation equations, are a common way of expressing physical laws. An example
is the equation of motion, Eq. (1.26), extended to a fixed amount of matter, which
can be written in the form
dPc T G
P + Ṗ
= Ṗ P
dt
where the momentum transfer, the first term on the right, is just the contact force,
Eq. (1.36), and the volume generation of momentum is due to gravity.
We will devote a large part of the remainder of this course to illustrating how
Eq. (4.6) can be applied in the solution of some typical engineering problems. We
use this equation in two ways to solve two distinct types of thermodynamic problems.
4.2.2 Steady State Problems
We know that equilibrium of a pure fluid is specified by the conditions
Q̇ = 0 Ẇ = 0
for every element of matter in the system. This implies, by virtue of Eq. (4.6), that, at
equilibrium, E is constant in time. However, the system energy can remain constant
in time even in the absence of equilibrium. We call such a situation a steady state,
by which we simply mean a situation unchanging in time. The difference between a
macroscopic equilibrium state, and a steady state is that, in the later, system properties
are not uniform in space, as they are in the former (recall Definition 2.1). Corollary
4.3 applies to this situation. It forms the basis for the solution of a class of problems
that can be addressed without knowledge of the energetic equation of state. In all
but the simplest cases, we make use of the fact that both Q̇ and Ẇ are additive
functions over the system surface area. In such problems, it is necessary to carefully
differentiate between the various systems that interact.
Example 4.1 On a cold winter day, the rate of heat transfer from the air in a
certain room is 5000 Btu/hr. The room is heated by a heating system, and the
warm air is circulated by a fan which develops 1/2 hp. What is the required rate
of heat transfer from the heating system in order to maintain constant room
conditions?
Solution.
Maintaining constant room conditions means that this is a steady-state prob-
lem. Therefore we apply Eq. (4.7) to the room air, which we take as our system
(see the sketch), and get Q̇ R = Ẇ R . We use Eq. (3.42) to calculate the heat
transfer to the room air
Q̇ R = Q̇ R/H + Q̇ R/O
so that
Q̇ R/H = Ẇ R − Q̇ R/O
Now according to the problem statement Q̇ R/O = −5000 Btu/hr. Moreover,

according to, Eq. (3.14), applied here
Ẇ R = −Ẇ F
where Ẇ F is 1/2 hp. Combining these facts in the equation for Q̇ R/H gives
Q̇ R/H = −(0.5 hp)(2544.5 Btu/hphr) − (−5000 Btu/hr) = 3728 Btu/hr

136 4 The First Law
We see that all the loss to the outside does not have to be replaced because of
the power developed by the fan.
Cyclic Processes
The designation steady state is also applied to systems that undergo cyclic processes
for which all system properties are periodic. Corollary 4.4, Eqs. (4.8) and (4.9),
applies to this case, which is seen to be the same as for the continuous process,
Eq. (4.7), except that here the result holds for the average power and heating (you
can also see this in Fig. 3.6 of Chap. 3).
4.2.3 Change of State Problems
Here we consider the two state class of problem that we encountered first in Chap. 2,
and integrate the energy equation, Eq. (4.6), over the process that connects the change
of state, namely from an initial time, t1 , to a final time, t2
(U2 − U1 ) + (K c2 − K c1 ) + (φc2 − φc1 ) = Q − W
The initial and final states are equilibrium states, K c1 = K c2 = 0, so as long as we

can neglect the change in potential energy of the mass center, we get
U2 − U1 = Q − W (4.11)
where U1 = Mu(v1 , T1 ) and U2 = Mu(v2 , T2 ) are the internal energies in the initial
and final states. On dividing this equation by the closed system mass, M, we obtain
the intensive form
u2 − u1 = q − w (4.12)
where q = Q/M and w = W/M are the heat transferred to, and the work done by
the system, per unit mass of substance. Neglect of the potential energy change of the
system mass center is justified in almost all thermodynamic problems of interest.
Accordingly, in a two-state closed system thermodynamic problem, one extensive
(mass or volume) and 5 intensive, quantities determine all the rest including work
and heat. In the usual specification, the mass, two properties of the initial and final
states, and the work done by the system, are given or can be inferred from the
problem statement. All other properties follow from the equations of state, and the
heat transferred to the system is calculable from Eq. (4.11).
Example 4.2 A mass, M, of a substance at pressure, p and temperature, T , is

heated slowly in a frictionless piston-cylinder arrangement until the temper-
ature is T ∗ . How much heat must be transferred to the substance in order to
accomplish this process?
Solution.
State 1 State 2
M1 = M T1 = T p2 = p1 M2 = M1
(piston (closed)
p1 = p equilibrium)
T2 = T ∗
W = M (p2 v2 − p1 v1 ) (isobaric
process)
Q = M (u2 − u1 ) + W (energy
equation)
Notice that we have written 8 equations corresponding to the information

given in the problem statement. As we discussed in the text 5 of these are
property specifications (4 intensive, 1 extensive), one is a work expression, one
is mass conservation, and one is Eq. (4.11). All completely specified problems
are of this type, however sometimes problems are incompletely specified and
intensive information extracted.
According to the description of the process given in the problem statement,
we can regard it as isobaric. Therefore it is a quasi-equilibrium process in
which the pressure is always the same, not just initially and finally, and the
work done by the substance is given by Eq. (3.34) as listed. Substituting the
expression for the work done by the substance into the energy equation gives
an expression for the heat transferred to the gas in an isobaric process
Q = M(u 2 − u 1 ) + M( p2 v2 − p1 v1 ) = M[(u 2 + p2 v2 ) − (u 1 + p1 v1 )]
(4.13)
In order to evaluate this, we need to know the equations of state that specify
u(v, T ) and p(v, T ). In the previous chapter we used an alternative expression
for the heat transfer in an isobaric process, obtained from Axiom 3.2 and
Eq. (3.59)
T2
Q∼Q=M c p dT
T1
In order to evaluate this we need to know the function c p ( p, T ). We

will investigate, in due course, the relation between these two expressions
for Q.
Enthalpy
The combination of properties that occurs in Eq. (4.13) is so common that we give
it a name, specific enthalpy
h = u + pv (4.14)
138 4 The First Law
This has an extensive counterpart which is obtained by multiplying by M
H = U + pV (4.15)
where H = Mh. According to Eq. (4.14) if we know the equations of state v( p, T )

and u[v( p, T ), T ], then we know also the enthalpic equation of state h( p, T ). In fact
knowledge of any two of these determines the third through Eq. (4.14).
4.3 The Energetic and Enthalpic Equations of State
The essential feature of Eq. (4.6), which we use to construct the energetic equation
of state, is that the difference in internal energy between any two equilibrium states,
is independent of the process by which the system arrived at the second from the
first. This allows us to write Eq. (4.11) as
U2 − U1 = Q − W = Q − W
wherein the second equality we use the work done and heat absorbed in any equi-
librium process that passes through the two states. Axioms 3.1 and 3.2 are our
justification for this step. Therefore the energetic equation of state, for a specific
substance, can be determined by using any convenient (even fictitious) process, and
once determined it can be used subsequently in the solution of problems involving
other processes of the same substance.
Using the equilibrium process defined by
v = v p (t) T = T p (t) p = p p (t) = p[v p (t), T p (t)]
to construct u = U/M = u p (t), we substitute the expressions for the equilibrium

power developed, Eq. (3.31), and the rate of heat absorption, Eq. (3.49), discussed in
the previous chapter, into Eq. (4.6), d E/dt = Q̇ − Ẇ, which on neglecting kinetic
and potential energies becomes
du p dT p dv p
= cv + (λv − p p ) (4.16)
dt dt dt
This is the differential equation of a space curve that describes the process; it lies in
the surface
u = u(v, T ) (4.17)
therefore from Eq. (2.18), du p /dt = ∂T u dT p /dt + ∂v u dv p /dt
∂T u = cv ∂v u = λ v − p (4.18)
4.3 The Energetic and Enthalpic Equations of State 139
These relations indicate that the energetic equation of state can be constructed from
experimental knowledge of the specific heats1 and pressure [see Eq. (2.23)]
T v
u(v, T ) = u O + cv (v O , TC ) dTC + [λv (vC , T ) − p(vC , T )] dvC (4.19)
TO vO
The condition on the equality of the mixed partial derivatives, Eq. (2.22), that insures
that u(v, T ) is a continuous surface here produces a relation between the derivatives
of the specific heats, that must be satisfied
∂v cv = ∂T λv − α/β (4.20)
Note that when this is true, it implies Eq. (3.58) as well. Therefore, the existence of
an energetic equation of state requires Q to be an interaction.
Enthalpic Equation of State
It is more convenient to use pressure rather than specific volume as an independent

variable for enthalpy. In order to do this we begin by substituting Eq. (3.50) instead
of Eq. (3.49) for Q̇ into the energy equation to get
du p dT p dp p dv p
= cp + λp − pp (4.21)
dt dt dt dt
Although we could express this in terms of the derivatives of p p and T p only, by
using the derivative of the mechanical equation of state, Eq. (3.51),
dv p dT p dp p
= vpα − vpβ
dt dt dt
to eliminate dv p /dt; it is more useful to proceed in an entirely different way.
By adding d( p p v p )/dt to each side of Eq. (4.21), using the chain rule for differ-
entiation on the right hand side, and combining terms on the left, we obtain
d dh p dT p dp p
(u p + p p v p ) = = cp + (λ p + v p ) (4.22)
dt dt dt dt
This is the differential equation for the specific enthalpy, Eq. (4.14); it lies in the
continuous surface which represents the enthalpic equation of state
h = h( p, T )
1 Here we write λv to keep the expression compact, however according to Eq. (3.50) it is equal to
(c p − cv )/vα, whenever c p has meaning and α = 0.
140 4 The First Law
its partial derivatives are given by
∂T h = c p ∂ph = λp + v (4.23)
it is calculable from knowledge of the specific heats and specific volume

T p
h( p, T ) = h O + c p ( p O , TC ) dTC + [λ p ( pC , T ) + v( pC , T )] dpC (4.24)
TO pO
and equality of the mixed partial derivatives requires
∂ p c p = ∂T λ p + vα (4.25)
4.3.1 Liquids and Solids
As we saw in Sect. 2.4.1, liquids and solids are incompressible v ∼ v O . Use of this
approximation implies that βλv << 12 , therefore, the second integral in Eq. (4.24)
can be evaluated [recall Eq. (3.53), λ p = −vβλv ]. Moreover for a limited range of
T , c p is essentially constant and in this case the first integral can be evaluated as
well, with the result
h ∼ h O + c p (T − TO ) + v O ( p − p O ) (4.26)
This result, together with v ∼ v O and h = u + pv, produces the energetic equation
of state
u ∼ u O + c p (T − TO ) (4.27)
However, since all physical quantities of interest depend only on differences in u or

h values, the choice of h O , u O , p O , or TO is immaterial. On subtracting the value of
either Eq. (4.26) or (4.27) in two states gives
u 2 − u 1 ∼ c p (T2 − T1 ) (4.28)
h 2 − h 1 ∼ c p (T2 − T1 ) + v O ( p2 − p1 ) (4.29)
Example 4.3 A 1.5 hp motor which has an efficiency of 92% is used to stir
5 lbm of water contained in a closed vessel, and initially at a temperature of
70 ◦ F. After 10 minutes the motor is switched off, and the process undergone
2 This inequality cannot be deduced from what we know at this point, but it follows easily from
results we will obtain in Chap. 5.
by the water was adiabatic. What is the final water temperature in this case?
You may regard the liquid water as an incompressible substance.
Solution.
State 1 State 2
M1 = Tf 1 = 70◦ F p2 = M2 = M1
5 lbm p1 = (incompressible) (closed)
(incomp- u2 = u1 − W/M
pressible) (energy eqn)
W = −{ẆM }(t2 − t1 )ηM (rotational work)

Q=0 (adiabatic)
No pressure information is given, but when the water is incompressible,

the mechanical and energetic equations of state do not depend on it. Therefore
we enter it among the given quantities as a blank. This tells us that we have
enough information to completely solve the problem, and that we should not
waste time trying to discover values for the pressure. Notice that, although the
initial and final pressures are the same, this is not an isobaric process. The work
done by the water is obtained by using Eq. (3.25), and {Ẇ M } = W M /(t2 − t1 ).
Substituting the given data this is
W = −(1.5 hp)(0.92)(550 ftlbf/hps)(600 s)/(778.2 ftlbf/Btu) = −585.2 Btu
Using Eq. (4.28) and (1.49), the final fahrenheit temperature is

T f 1 − W/Mc p
Tf2 =
70 ◦ F − (−585.2 Btu)/[(5lbm)(1 Btu/lbm R)(1 R/ ◦ F)]
which is 187 ◦ F.
Note that this example could not have been solved by the methods discussed in the
previous chapter. The energy equation, Eq. (4.12), is a crucial new element.
Vaporizable Liquids
As we saw in Chap. 2 real liquids are incompressible and vaporizable. Therefore,
we modify the path of integration in Eq. (4.24), as we did previously in Chap. 2 (see
Fig. 2.7). Making the same incompressibility approximation as before, we now get
h( p, T ) ∼ h f (T ) + v f (T )[ p − ps (T )] p ≥ ps (T ) (4.30)
where
T ps (T )
h f (T ) = h O + c p [ ps (TC ), TC ] dTC + {λ p [ pC , Ts ( pC )] + v̂ f ( pC )} dpC
TO pO
142 4 The First Law
is the enthalpy of the boiling liquid. The energetic equation of state is obtained by
using Eq. (4.14), h = u + pv, and v ∼ v f . It is
u(v, T ) ∼ u f (T ) p ≥ ps (T ) (4.31)
so that the liquid internal energy depends only on temperature. The quantity u f is
the specific internal energy of the boiling liquid
u f = h f − ps v f
Values of h and u are calculated, using Eqs. (4.30) and (4.31), for various substances
and tabulated. These calculations require choices for h O , p O , and TO , however as
before (see Sect. 4.3.1) property differences do not depend on them. The tables are
used to solve problems as the following example shows.
Example 4.4 A 1.5 hp motor which has an efficiency of 92% is used to stir
5 lbm of water contained in an open vessel, initially at a temperature of 70 ◦ F.
When the motor is switched off and the water has come to rest its temperature
has just reached the boiling point. Assuming that no water has boiled away,
that the process is adiabatic, and that the atmospheric pressure is 14.7 psi, how
long did the motor run?
Solution.
State 1 State 2
M1 = Tf 1 = 70◦ F p2 = 14.7 psi M2 = M1
5 lbm p1 = 14.7 psi (open to (nothing
(open to atmosphere) boiled away)
atmosphere) x2 = 0
W = Q − M1 (u2 − u1 ) (energy equation)

Q=0 (adiabatic)
Although the initial and final pressures are the same, this is not an isobaric
process, the work done by the water is obtained by using the energy equation.
The initial and final internal energies can be found in the property tables for
water as the saturated liquid values, Eq. (4.31); the initial value requires a
simple interpolation, but the final value does not. The work done by the water
is then
W = −[(180.11 − 38.05) Btu/lbm](5 lbm) = −710.3 Btu
It is interesting to note that since c p of water is nearly 1 Btu/lbm R (recall

Sect. 3.2.2), the energy difference obtained by using Eq. (4.28) as an approx-
imation is
u 2 − u 1 = (211.99 − 70) Btu/lbm = 141.99 Btu/lbm
which is within 0.05% of the tabulated value. The running time is obtained from
Eq. (3.3), {Ẇ M } = W M /(t2 − t1 ), together with Eq. (3.25), W = −η M W M ,
⎧
⎨ (−W )/(η M {Ẇ M })
t2 − t1 = (710.1 Btu)/[0.92(1.5 hp)(2545 Btu/hp hr)]
⎩
0.202 hr
which is 12.1 min.
This example also could not have been solved by the methods of the previous chapter.
4.3.2 Gases
Equations (4.19) and (4.24) are also applied to gases. As with the mechanical equation
of state studied in Chap. 2, simplifications occur for ideal gases, and to an even greater
extent when the specific heats, cv and c p , are constant, a material we have called a
perfect gas [see Sect. 4.1].
Ideal Gases
In addition to his fundamental experiments relating to the equivalence of heat and
work, J. P. Joule also made an important discovery about the character of the energetic
equation of state for ideal gases. His original experiment, depicted in Fig. 4.2, allowed
a gas to freely expand, from volume A into A + B, while in contact with a calorimeter.
After adjusting the valves to evacuate the volume B, the expansion was repeated, and
the entire process was repeated for several lower starting pressures. At every stage
of the experiment, Joule noted that the temperature of the calorimetric substance
was unchanged. This indicated that the expansions were occurring adiabatically.
Fig. 4.2 In the free expansion experiment valve 1 is closed and 2 is opened while the gas is pumped
out of part B of the volume. Then valve 2 is closed, and valve 1 is opened to allow the gas in part
A to expand into B. The thermometer reading is noted at the end of the expansion process; if no
change has occurred, there has been no heat transfer between the gas and the water. Valve 1 is then
closed and valve 2 opened, and the entire process is repeated at a lower starting pressure
144 4 The First Law
Furthermore since there was no work done in this free expansion processes [recall
Eq. (3.27)], the energy equation, Eq. (4.12), required the internal energy at each stage
to be unchanged, u 1 (T, v1 ) = u 2 (T, v2 ) = · · · = u n (T, vn ). Joule reasoned that this
was only possible if the energetic equation of state was independent of the specific
volume
u = u(T ) (4.32)
This realization has several consequences. Using Eq. (4.32) in Eq. (4.18) shows
that λv = p. Substituting this into Eq. (3.54), λv = (c p − cv )/vα gives
c p − cv = pvα (4.33)
and Eq. (4.19) can be written as a single integral

T
u(T ) = u O + cv (TC ) dTC (4.34)
TO
Substituting the ideal gas equation of state into Eq. (4.33) along with the ideal gas
value 1/T for α shows that the specific heats are simply related
c p = cv + R (4.35)
On putting this into Eq. (4.34) and using the definition, Eq. (4.14), h = u + pv, along
with the ideal gas equation of state, we obtain
T
h(T ) = h O + c p (TC ) dTC (4.36)
TO
Therefore the ideal gas enthalpy, like the internal energy, depends on temperature
only.
These results indicate that all thermodynamic information relating to a specific
ideal gas is contained in a single function, say c p (T ), along with the value of the gas
constant R and ideal gas equation of state. For some common gases, tables of internal
energy and enthalpy as functions of temperature have been prepared from measured
values of this specific heat, thus saving students the necessity of evaluating the
integrals in Eqs. (4.32) and (4.36). Creating these tables requires choices for u O , h O
and TO , however as before with liquids, property differences are independent of these
values. For example, from Eq. (4.36)
T2 T1 T2
h2 − h1 = c p (T ) dT − c p (T ) dT = c p (T ) dT
TO TO T1
Example 4.5 Air contained by a frictionless piston in a 3 ft3 cylinder at pres-

sure, 100 psi and temperature, 360 ◦ F, is heated slowly. If the piston compresses
a linear spring as it rises, the heating continues until the volume increases by
30%, and the final pressure is 180 psi, what is the final temperature of the air,
and how much heat was transferred to it?
Solution.
State 1 State 2
V1 = Tf 1 = 360◦ F p2 = 180 psi M2 = M1
3 ft3 (closed)
p1 = 100 psi V2 = 1.3 V1
(stated)
W = M (p1 + p2 )(v2 − v1 )/2 (piston quasi-

equilibrium)
Q = M (u2 − u1 ) + W (energy
equation)
According to the description, this is a quasi-equilibrium process of a spring

loaded piston. Therefore the work done by the air is given by Eq. (3.35) as
listed above. The information required for the calculation is all given
W = [(100 + 180) psi][(3.9 − 3.0) ft 3 ](144 in2 /ft2 )/2 = 18144 ftlbf
or W = 23.32 Btu. The final temperature of the air is determined, as usual, by

the ideal gas equation of state applied at both end states
p1 V1 /M1 T1 = p2 V2 /M2 T2
On making use of the given data this is
T2 = (180/100)(3.9/3)(360 + 459.67) R = 1918 R (T f 2 = 1458 ◦ F)
Values for internal energy are obtained from the ideal gas tables for air with
a knowledge of temperature. For example the energy in state 1 is found by
interpolating in the table (using the chord line connecting the points with
temperatures 350 and 375)

29.327 + [(33.740 − 29.327)/(375 − 350)](360 − 350)
u1 =
31.09 Btu/lbm
146 4 The First Law
and the energy in state 2 is found by interpolating in the table (using the chord
line connecting the points with temperatures 1450 and 1500)

241.84 + [(252.29 − 241.84)/(1500 − 1450)](1458 − 1450)
u2 =
243.24 Btu/lbm
In order to calculate the heat transfer we need the mass of air. This is obtained
from the equation of state applied to state 1

(100 psi)(3.0 ft3 )/(0.3704 psi ft3 /lbm R)(819.67 R)
M1 = p1 V1 /RT1 =
0.9881 lbm
Collecting all this into the expression for Q
Q = (0.9881 lbm)[(243.24 − 31.09) Btu/lbm] + 23.32 Btu
which gives 232.9 Btu for the amount of heat transfer to the air.
For some purposes (for example, Sect. 3.3.5), the ratio of specific heats
k = c p /cv
is a useful quantity. In terms of this ratio, we get directly from Eq. (4.35)
1
cv = R (4.37)
k−1
k
cp = R (4.38)
k−1
Perfect Gases
It is reasonable to expect that over a limited range of temperature, the specific heats
only vary slightly. Moreover our simple kinetic model, Eq. (4.5), along with Eq. (1.41)
and Eq. (2.39), produces
U = (3/2) pV = (3/2)M RT
From this we get immediately cv = 3R/2 and then on using Eq. (4.35) we have
c p = 5R/2 so that k = 5/3 = 1.667. These results apply to gases composed of atoms
having only the 3 degrees of freedom corresponding to motion in the three spatial
directions. Argon, Helium, and Neon are examples of such gases and their measured
specific heat values are very close to these around normal room temperatures. If
we regard Oxygen and Nitrogen (the major components of air) as composed of
molecules containing two atoms a fixed distance apart, then they have two additional
degrees of freedom corresponding to rotational motion in two directions (there is little
energy associated with spin about the line connecting the two atoms). Therefore for
these gases cv = 5R/2, c p = 7R/2, and k = 7/5 = 1.4. These values are also in
agreement with measurements around room temperature3
For constant specific heat, performing the integrations in Eqs. (4.34) and (4.36).
This gives,
u = u O + cv (T − TO ) h = h O + c p (T − TO ) (4.39)
Notice that the values of u O , h O and TO are completely arbitrary, but that physical
quantities such as Q and W do not depend on what choice you make. The equations
of state are always used in difference form [for example see Eq. (4.12)]. Taking
differences of Eqs. (4.39) produces
u 2 − u 1 = cv (T2 − T1 ) h 2 − h 1 = c p (T2 − T1 ) (4.40)
This point was discussed previously in Sect. 4.3.1 in connection with liquids and
solids, and in connection with ideal gases as well.
The two equations of state in Eq. (4.39) may be said to define a perfect gas.
Together they imply the ideal gas equation of state, as can be seen by subtracting the
first from the second and using the result in Eq. (4.14), h = u + pv. Thus a perfect
gas is an ideal gas that has constant specific heats. A particular perfect gas is therefore
described simply by two numbers R and k.
Example 4.6 Air contained in a 3 ft3 cylinder at pressure, 100 psi and temper-
ature, 360 ◦ F, is heated slowly in a cylinder. The piston compresses a linear
spring as it rises, the heating continues until the volume increases by 30%, and
the final pressure is 180 psi. Assuming that air is a perfect gas with specific heat
ratio 1.4, calculate the final temperature, and the amount of heat transferred to
the air?
Solution.
State 1 State 2
V1 = Tf 1 = 360◦ F p2 = 180 psi M2 = M1
3 ft3 (closed)
p1 = 100 psi V2 = 1.3 V1
(stated)
W = M (p1 + p2 )(v2 − v1 )/2 (quasi-equilibrium

spring deflection)
Q = M (u2 − u1 ) + W (energy
equation)
3 The temperature variation of the specific heats of these gases, and consequently of air, occurs
because the bond between the atoms is not completely rigid so that at higher temperatures there
are additional degrees of freedom, related to vibration, that contribute increasingly to the internal
energy.
148 4 The First Law
Fig. 4.3 The path of integration for enthalpy in the vapor region is the constant pressure curve
extending from the saturated vapor curve to the desired temperature, as shown in these two views
The difference between this example and the previous one is that here we
treat the air as a perfect gas. Consequently, the work done by the air, its mass,
and final temperature are the same as previously. Here we use Eq. (4.40) for
u 2 − u 1 with cv found in the table as 0.171 Btu/lbm. Thus we find for Q

(0.9881 lbm)(0.171 Btu/lbm R)(1458 − 360) R + 23.32 Btu
Q=
208.84 Btu
The 10% difference in the result for Q between this example and the pre-
vious one is due to the neglect here of the variation in cv with temperature.
Condensible Gases
It is usual in this regime to use p and T as independent variables; as we have already
noted, this is the way the steam tables are constructed. Then if we use the enthalpy,
Eq. (4.24), and integrate from the saturation curve up to the desired temperature at
constant pressure, see Fig. 4.3,
T
h( p, T ) = h g + c p ( p, TC ) dTC T ≥ Ts ( p)
Ts ( p)
This gives h in terms of the measured values of c p and the value of the enthalpy of
saturated vapor, h g . The latter cannot be chosen arbitrarily; it is related to the enthalpy
of the saturated liquid, h f . Indeed the two states, saturated liquid and saturated vapor
have the same pressure, so in a process for which a mass M of liquid absorbs heat and
evaporates, the amount of heat transfer is given by Eq. (4.13), in terms of enthalpy
Q = M(h g − h f ) (4.41)
In other words, h g is determined from h f by measuring q, the amount of heat transfer

required to evaporate a unit mass of liquid.
When the enthalpy is known, the internal energy follows from Eq. (4.14) using
the mechanical equation of state in the form v = v( p, T ). Once a table of values of
h and u is built in this way for a specific substance, we can subsequently use it to
solve a variety of problems.
Example 4.7 Steam contained in a 3 ft3 cylinder at pressure, 100 psi and
temperature, 360 ◦ F, is heated slowly in a cylinder. The piston compresses a
linear spring as it rises, the heating continues until the volume increases by
30%, and the final pressure is 180 psi. What is the final temperature, and the
amount of heat transferred to the steam?
Solution.
State 1 State 2
V1 = Tf 1 = 360◦ F p2 = 180 psi M2 = M1
3 ft3 (closed)
p1 = 100 psi V2 = 1.3 V1
(stated)
W = M (p1 + p2 )(v2 − v1 )/2 (quasi-equilibrium

spring deflection)
Q = M (u2 − u1 ) + W (energy
equation)
The difference between this example and the previous is that here we cannot
treat the steam as an ideal gas. Consequently only the work done by the steam
is the same as previously. The quantities v1 and u 1 are given in the table
v1 = 4.661 ft3 /lbm u 1 = 1119.3 Btu/lbm
From this we can determine the mass, M1 = V1 /v1 = 0.6436 lbm. The final
state is specified by the intensive variables, pressure, and specific volume; the
later is
v2 = V2 /M2 = V2 /M1 = 6.0593 ft3 /lbm
We find T2 and u 2 by interpolation using the chord line at 180 psi

1300 + [(1500 − 1300)/(6.4704 − 5.8014)](6.0593 − 5.8014)
Tf2 =
1377 ◦ F
150 4 The First Law
and

1497.0 + [(1585.7 − 1497.0)/(6.4704 − 5.8014)](6.0593 − 5.8014)
u2 =
1531.19 Btu/lbm
The final temperature here differs by about 5% from the value calculated
previously because the steam is not an ideal gas in state 1. The heat transfer is
calculated from the assembled data
Q = (0.6436 lbm)[(1531.19 − 1119.3)Btu/lbm] + 23.32 Btu
or 288.4 Btu. This result is a third greater than the ideal gas value calculated
previously, which indicates that non-ideality can be an important practical
effect.
4.3.3 Liquid–Vapor Equilibrium
For processes that either begin or end in a state corresponding to an equilibrium

between phases, we need appropriate expressions for the energetic and enthalpic
equations of state. Since U and H are extensive we can write, just as we did for
volume in Sect. 2.4.3,
U = U f + Ug = M f u f + M g u g
H = H f + Hg = M f h f + Mg h g
where M f , U f , H f , u f , and h f apply to the liquid phase and Mg , Ug , Hg , u g , and

h g are the corresponding quantities for the vapor phase. Dividing by the total mass,
M = M f + Mg gives the equations of state in the form
u = (1 − x)u f + xu g = u f + x(u g − u f ) (4.42)

h = (1 − x)h f + xh g = h f + x(h g − h f ) (4.43)
where the quality, x = Mg /M, is the same quantity that appeared previously in
Eq. (2.47). Of course in order to use these equations to determine u and h, the liquid,
u f , h f and vapor u g , h g properties must be known; either from knowledge of the
saturation temperature or saturation pressure.4
4 Recallthat in phase equilibrium pressure and temperature are not independent, and knowledge of
one of them implies the other.
Example 4.8 One and a half pounds of water at 200 psi and 90% quality is
contained in part of a vessel; the other part of the vessel is evacuated. When
the dividing partition is slid aside, the water freely expands to fill the entire
vessel, and comes to rest with a pressure of 40 psi. If the process is adiabatic,
what is the final temperature? How many times greater is the final volume than
the initial?
Solution.
State 1 State 2
M1 = p1 = 200 psi p2 = 40 psi M2 = M1
1.5 lbm u2 = u1 (fixed mass)
x1 = 90% (energy
equation)
W =0 (free
expansion)
Q=0 (adiabatic)
The initial state is a liquid–vapor equilibrium, so the quantities v1 and u 1

can be determined using Eqs. (2.47) and (4.42) along with the data provided

[0.01839 + 0.9(2.2873 − .01839)] ft3 /lbm
v1 =
2.060 ft3 /lbm

[354.83 + 0.9(1113.7 − 354.83)] Btu/lbm
u1 =
1037.81 Btu/lbm
In addition, the initial Fahrenheit temperature T f 1 is 381.8 ◦ F, and the volume

can be determined from
V1 = v1 M1 = 3.09 ft3
The final state is specified by the intensive variables, pressure, and specific
internal energy. Since at the final pressure of 40 psi, u g is greater than u 2 ,
the final state is also a liquid–vapor equilibrium. Consequently the final tem-
perature is the saturation temperature, 267.3 ◦ F. The final specific volume is
calculated by first determining the quality from u 2

(1037.81 − 236.01)/(1092.1 − 236.01)
x2 =
0.9366
From this v2 follows using Eq. (2.45) applied to state 2

152 4 The First Law

[0.01715 + 0.9366(10.497 − .01715)] ft3 /lbm
v2 =
9.983 ft3 /lbm
The volume ratio is the same as the specific volume ratio for a closed system,
in this case it is 4.77.
4.4 The Open System
Although the energy equation, Eq. (4.6) applies to a fixed quantity of matter, in many
applications it is more convenient to analyze a fluid system whose mass can vary,
an open system. We can derive equations applicable to an interesting class of open
systems by subdividing a closed system into several parts, which we denote by the
subscripts, V , for the system, and called a control volume, and j (1 ≤ j ≤ N ), for
a set of N feed pipes. In each feed pipe near the entrance to the control volume the
fluid is in a macroscopic equilibrium state, however the state varies from one pipe
to another, and from place to place within the volume V . A typical configuration
is shown in Fig. 4.4. For this single feed pipe geometry the mass and energy of the
closed system can be written as,
M = MV + M j
E = EV + E j
Differentiating these, noting that the mass is constant, gives
dM d MV dMj
= + =0 (4.44)
dt dt dt
and using Eq. (4.6) for the increase in energy gives,
dE d EV dEj
= + = Q̇ − Ẇ (4.45)
dt dt dt
Now the amount of mass in the feed pipe, Eq. (1.23)
M j = ρ j Vj
changes because its volume changes
dMj dVj
= ρj (4.46)
dt dt
and the rate of increase of the feedpipe volume, as can be seen in Fig. 4.4 (this is a
special case of Fig. 3.5, see also Sect. 3.2.5), is simply
4.4 The Open System 153
Fig. 4.4 Model for the

development of an energy
equation in which the mass
can vary
dVj
= Ajvj · nj
dt
where n j is an outward pointing unit vector from V at the feedpipe, and A j is the
pipe cross-sectional area. Defining the volume rate of flow by
V̇ j = v j A j (4.47)
we can write
dVj
= V̇ j (e j · n j ) (4.48)
dt
where e j is a unit vector in the direction of the flow velocity. The dot product appear-
ing here is
−1 for flow into the volume, V
e j · n j = sign j = (4.49)
1 for flow out of the volume, V
Combining Eqs. (4.46), (4.48), and (4.49), and substituting into Eq. (4.44) produces
d MV
= −sign j ρ j V̇ j = −sign j Ṁ j (4.50)
dt
where the mass rate of flow, Ṁ j is defined as
Ṁ j = ρ j V̇ j = ρ j v j A j = v j A j /v j (4.51)
Physically Eq. (4.50) expresses the conservation of mass, namely that the rate of
increase of mass in the control volume is equal to the rate at which mass flows into
the volume.
The amount of energy in the fluid in the feedpipe is
E j = e j M j = (u j + v2j /2 + gz j )M j
154 4 The First Law
so that on differentiating
dEj dMj
= ej = sign j e j Ṁ j (4.52)
dt dt
In addition
Q̇ = Q̇ V + Q̇ j (4.53)
Ẇ = ẆV + Ẇ j (4.54)
where the power developed by the fluid in the feedpipe as a result of the motion of
one of its surfaces is assumed to be the equilibrium power, Eq. (3.31)
dVj
Ẇ j ∼ Ẇ j = p j = sign j p j V̇ j = sign j p j v j Ṁ j (4.55)
dt
and Q̇ j = 0 because there is no significant area available for heat transfer in the
feedpipe. The work term is called flow work, and physically it represents the amount
of work required to add that amount of mass to the control volume. Substituting
Eqs. (4.52)–(4.55) into Eq. (4.45) produces
d EV
= −sign j (e j + p j v j ) Ṁ j + Q̇ V − ẆV (4.56)
dt
The quantity multiplying the mass rate of flow is
e j + p j v j = h j + v2j /2 + gz j (4.57)
The appearance of enthalpy in this expression is a consequence of the flow work,

and gives the interpretation of flow energy to the enthalpy. Equation (4.56) is an
expression of conservation of energy for the control volume.
For a system with N feedpipes the equations for mass and energy conservation
generalize to (and dropping the system subscript for simplicity)
dM N
=− sign j Ṁ j (4.58)
dt j=1
dE N
=− sign j (e j + p j v j ) Ṁ j + Q̇ − Ẇ (4.59)
dt j=1
Accordingly we can write balance equations that specify the evolution of mass,
momentum, and energy within a volume as [recall Eq. (4.10)]
⎫
d M/dt = ṀC + ṀT + ṀG ⎬
C T G
P + Ṗ
dPc /dt = Ṗ P + Ṗ P (4.60)
⎭
d E/dt = Ė C + Ė T + Ė G
Here in addition to the familiar terms due to transfer and generation, there is a new
term due to convection. Thus when matter enters a volume it increases not only the
mass inside, but also the momentum and energy because it carries those properties
as well. The terms ṀT and ṀG , which are written in Eq. (4.60) for the sake of
symmetry are identically zero; the first because the boundaries of the volume are
impermeable, and the second because of mass conservation.
Although in general, Eq. (4.60) must be integrated, either analytically or numer-
ically, there is a useful class of problems that we consider in thermodynamics, for
which the mass and energy balances can be obtained algebraically.
4.4.1 Steady Flow Problems

In a steady flow problem, no extensive property characterizing the control volume,
V , changes with time. Thus both MV and E V are constant, and this defines, as it did in
Sect. 4.2.2, a steady state problem. Here however, we consider systems that exchange
mass with their surroundings whereas previously the systems always contained the
same matter, not just the same amount of matter. If the system has one inlet ( j = 1),
and one outlet ( j = 2), Eq. (4.58) reduces to
Ṁ1 = Ṁ2 (4.61)
and denoting the common value by Ṁ, Eq. (4.61) becomes
Ṁ(h 2 + v22 /2 + gz 2 ) − Ṁ(h 1 + v12 /2 + gz 1 ) = Q̇ V − ẆV (4.62)
On writing q = Q̇ V / Ṁ, and w = ẆV / Ṁ for the heat transfer and work done per
unit mass of fluid flow, we obtain the intensive form
(h 2 + v22 /2 + gz 2 ) − (h 1 + v12 /2 + gz 1 ) = q − w (4.63)
Interestingly, Eq. (4.63) looks very much like Eq. (4.12). Indeed the solution of
problems of steady flow devices with one inlet and one outlet, exactly parallels
what we have already done for two state problems of closed systems. However the
appearance of velocity in Eq. (4.63) couples the thermal and mechanical aspects of the
flow, and consequently complicates any attempt at solution. Therefore simplifications
to Eqs. (4.61) and (4.63) are introduced. These simplifications, which are based on
whether the fluid is a liquid or gas, are useful in applications, so we will develop
them here, and subsequently provide the examples.
Gases
In most problems involving the flow of gases, the kinetic energy difference is neg-
ligible compared with the enthalpy difference. Moreover in virtually all gas flows
the height differential is small enough so that the potential energy difference is also
negligible compared with the enthalpy difference. In such cases we can drop these
terms and use, in place of Eq. (4.63), the simpler form of the energy equation.
156 4 The First Law
h2 − h1 = q − w (4.64)
Thus the relationship between two state closed system problems, and two feedpipe
open system steady state problems for gases is simple and can be summarized as
follows:
Closed System Open System

M2 = M1 =⇒ Ṁ2 = Ṁ1
u 2 − u 1 = q − w =⇒ h 2 − h 1 = q − w
q = Q/M =⇒ q = Q̇/ Ṁ
w = W/M =⇒ w = Ẇ / Ṁ
v = V /M =⇒ v = V̇ / Ṁ
Moreover for piston cylinder problems, z = V /A =⇒ ż ≡ v = V̇ /A.

Liquids
Because of the much larger density of liquids relative to gases, the kinetic and poten-
tial energy terms usually cannot be neglected for these flows. However since liquids
are highly incompressible
v1 = v( p1 , T1 ) ∼ v( p2 , T2 ) = v2
Eq. (4.61), Ṁ2 = V̇2 /v2 = V̇1 /v1 = Ṁ1 , is approximated by
V̇1 ∼ V̇2 (4.65)
and this equation is simpler to use with Eq. (4.63) because it doesn’t involve state
properties. However, in these steady flow problems five new quantities, area A, and
flow speed v, in each state, and the height difference z 2 − z 1 between the states, in
addition to the usual, Ṁ, V̇ , p, v, T , and h are involved in the solution. Therefore
there are 19 variables, including work and heat transfer (8 · 2 + 1 + 2), instead of
the usual 14 for closed systems. Since an additional two equations [Eq. (4.47) for the
inlet and the outlet] relate the variables, we need to write 11 equations (instead of
8) resulting from the problem specification in order to determine all the remaining
variables.
4.4.2 Steady Flow Devices
Many different types of fluid machinery operate in a steady state, and can be analyzed
as open systems. Some of these important mechanical engineering applications are
illustrated in the following subsections.
Turbines, Compressors, and Pumps

A turbine is a device whose purpose is to produce mechanical work from the internal
energy of a fluid flow. There are gas turbines and liquid turbines. The former take in
gases at high pressure, and exhaust them at lower pressure, they convert the internal
energy of the gas into work. The later operate similarly with liquids. Conversely,
a compressor uses work to raise the pressure of a gas while a pump uses work to
increase the pressure in a liquid. The following examples illustrate the solution of
problems involving these devices.
Example 4.9 The turbine in a dentist’s drill operates with a supply of air at
its inlet of 50 psi and 80 ◦ F, exhausts the air at 15 psi and 0 ◦ F, has an inlet
area of 0.085 in2 , and a ratio of exit to inlet area of 4.5. If the inlet volume
rate of flow is 0.04 ft3 /s, and the process is adiabatic, how much horsepower
is developed? What is the airspeed at the inlet and the outlet? If the rotational
speed of the turbine is 350,000 rpm, what is the torque output? Assume that
air under these conditions behaves like a perfect gas.
Solution.
State 1 State 2
Tf 1 = 80◦ F p2 = 15 psi
V̇1 = Ṁ2 = Ṁ1
0.04 ft3 /s p1 = 50 psi Tf 2 = 0◦ F (steady
flow)
A1 = 0.085 in2 A2 = 4.5A1
stated
Ẇ = Ṁ1 (h1 − h2 ) + Q̇ (energy equation)

Q̇ = 0 (adiabatic)
The inlet airspeed is obtained using Eq. (4.47), V̇1 = v1 A1
v1 = 0.04 ft3 /s · 144 in2 /ft2 /0.085 in2 = 68 ft/s
The outlet airspeed is related to this by the steady flow condition
v2 = (A1 /A2 )(v2 /v1 )v1 = (A1 /A2 )( p1 / p2 )(T2 /T1 )v1
and on using the information provided we get 43 ft/s. The initial specific volume
is obtained from the equation of state
v1 = (53.34 ft lbf/lbm R · 539.67 R)/(50 lbf/in2 · 144 in2 /ft2 ) = 4 ft3 /lbm
and subsequently the mass rate of flow, Ṁ1 , from Eq. (4.51) is 0.01 lbm/s. The
power output is then, using the tabulated value for c p for air and Q̇ = 0
158 4 The First Law

0.01 lbm/s · 0.240 Btu/lbm R · (80 − 0) R · 1.415 hp s/Btu
Ẇ =
0.272 hp
Finally the torque output can be calculated from the rotational power expres-
sion, Eq. (3.28), Ẇ = Mk ω,

(0.272 hp · 550 ft lbf/s hp · 12 in/ft)/36652 rad/s
Mk =
0.0485 in lbf
Example 4.10 A compressor in an air conditioning system operates with 60 ◦ F,

saturated vapor R-12 at its inlet, and 110 psi, 100 ◦ F at its outlet. If there is
heat transfer to the surrounding air in the amount 3.4 Btu for each lbm of R-12
flowing through the compressor, what is the required amount of work per unit
mass flowing to drive it?
Solution.
State 1 State 2
Tf 1 = 60◦ F p2 = 110 psi
Ṁ2 = Ṁ1
x1 = 1 Tf 2 = 100◦ F (steady
flow)
w = q + (h1 − h2 ) (energy
equation)
q = −3.4 Btu/lbm
In this problem, only 7 of the required 10 pieces of information are given.

Consequently, we cannot determine all the variables, in particular, we cannot
know the areas, velocities, mass rate of flow and volume rate of flow. Also note
that q is the negative of the heat transfer to the air from the compressor, and w
is the negative of w M , the work done by the motor that drives the compressor.
In order to calculate w we need to first obtain the initial and final enthalpy.
The initial enthalpy is that of saturated vapor at the stated temperature, thus
from the table of R-12 properties h 1 = 84.012 Btu/lbm. The final enthalpy is
obtained by interpolating the tabulated values using the chord line at 100 ◦ F

88.353 + (89.287 − 88.353)(110 − 120)/(100 − 120)
h2 =
88.820 Btu/lbm
We then obtain from the energy equation

w = [(84.012 − 88.820) − 3.4] Btu/lbm
The work required to drive the compressor is the negative of w; 8.2 Btu for
each lbm of R-12.
Example 4.11 At the Hoover Dam on the Colorado river the turbines that are
used to produce hydroelectric power are located 712 ft below the surface of
Lake Mead. At the inlet of the turbine flume 2 ft below the surface, the pressure
is approximately atmospheric and the temperature is 60 ◦ F, while at the outlet,
located at the same height as the turbine, the area, temperature, and pressure
are the same as the inlet. Assume the process is adiabatic, and calculate the
work done per pound of water that flows? What volume flow rate is required
to produce 1.6 MW of power?
Solution.
State 1 State 2
Ṁ1 = Ẇ /[(h1 − h2 )+ p1 = pa p2 = pa
(v12 − v22 )/2+ (stated) V̇2 = V̇1
g(z1 − z2 )] Tf 1 = T2 = T1 (steady
(energy eqn) 60◦ F (stated) flow)
z1 = z2 + 710 ft A2 = A1
(stated) (stated)
Ẇ = 1.6 MW
Q̇ = 0 (adiabatic)
In this problem only, 10 of the required 11 pieces of information are given.

Thus, as in the previous problem, we cannot obtain values for all the variables,
specifically the areas and velocities. However since both pressure and tem-
perature are equal in the two states, then h 2 = h 1 [see Eq. (4.29)], and since
the two areas are equal, the steady-state condition produces v2 = v1 . All this
means that the mass rate of flow is expressed by (using the fact that Q̇ = 0)
⎧
⎪
⎪ Ẇ /[g(z 1 − z 2 )]
⎨
(1.6 · 103 kW)(738 ft lbf/kW s)(32.17 lbm ft/lbf s2 )/
Ṁ1 =
⎪
⎪ [(32.17ft/s2 )(710 ft)]
⎩
1660 lbm/s
The volume rate of flow is then
V̇1 = v Ṁ1 = (0.016033 ft3 /lbm)(1660 lbm/s)
which is 26.7 ft3 /s or 11940 gal/min. Note the use of conversion factors here.
160 4 The First Law
Simple Electrical Devices

Electric motors, generators, and heaters require a flow of electric current for their
operation. As you learned in physics, electric charges exert forces on one another,
and these forces develop power when the charges move (a moving charge is a current
of electricity). Now any of these devices in an electric circuit constitutes an open
thermodynamic system like the ones we have been discussing, however in such
systems the convected energy terms that we have previously discussed, which depend
on the mass rate of flow of the electrons, are negligible compared with the convected
energy associated with the charge rate of flow, the electric current. The form of these
two kinds of terms is similar, as you can see from the following comparison for
steady flow.
Ṁ[(h 2 + v22 /2 + gz 2 ) − (h 1 + v12 /2 + gz 1 )], Ċ[φe2 − φe1 ]
We will write this electrical power term as
Ẇe = Ċ[φe2 − φe1 ] = ±I φ (4.66)
where I is the current flow in the circuit measured in amperes (a), φ is the voltage
drop across the circuit measured in volts (v), and Ẇe is the electrical power devel-
oped by the control volume, negative for motors and heaters [use the minus sign in
Eq. (4.66)], and positive for generators [use the plus sign in Eq. (4.66)]. The product
of a volt and an ampere is a watt, and these electrical SI units are used also with
English System units; the appropriate conversion relation is
1 kW = 738 ft lbf/s = 56.9 Btu/min (4.67)
A thermodynamic system that contains such an electrical device thus is governed

by an energy equation which is a generalization of Eq. (4.59)
dE N
=− sign j (e j + p j v j ) Ṁ j + Q̇ − Ẇ − Ẇe (4.68)
dt j=1
or for a steady flow device with one inlet and one outlet
Ṁ(h 2 + v22 /2 + gz 2 ) − Ṁ(h 1 + v12 /2 + gz 1 ) = Q̇ − Ẇ − Ẇe (4.69)
where Ẇe is calculated from Eq. (4.66). We recognize a system as one in which
electrical power is developed by noting whether or not electric wires cross the system
boundary and carry a current. If the answer to this question is yes, we must include
this new term.
Equation (4.66) applies to alternating current as well as direct. In the former case,
the current and voltage values used are the rms (root mean square) values, and the
power is the average power. This is the electrical equivalent of the cyclic, steady-state
relationship we noted before in Sect. 4.3.1.
Example 4.12 Air flows in a uniform duct that contains a heater that draws 16
a from a 440 v line. If the air enters at 70 ◦ F, 20 psi, with a speed of 3 ft/s, exits
at 16 psi with a mass flow rate of 30 lbm/min, and there is no heat transfer to
the duct walls, what is the final air temperature? Treat the air as a perfect gas
under these conditions.
Solution.
State 2
State 1
p2 = 16 psi
p1 = 20 psi
h2 = h1 − Ṁ2 =
Ṁ1 = Ṁ2 Tf 1 = 70◦ F
Ẇe /Ṁ2 30 lbm/min
(steady
(energy
flow)
equation)
v1 = 3 ft/s
A2 = A1
uniform duct
Ẇ = 0 (no moving surface)

Q̇ = 0 (no heat transfer)
Ẇe = −16 a · 440 v (electrical work)
The energy equation is, using Eq. (4.40), h 2 − h 1 = c p (T2 − T1 ), and

Eq. (1.49)
T f 2 = T f 1 − Ẇe /(c p Ṁ2 )
Using the values provided then gives
T f 2 = 70 ◦ F + (7.04 kW · 56.9 Btu/kW min)/(0.24 Btu/lbm R · 30 lbm/min · 1 R/ ◦ F)
for a final temperature of 125.6 ◦ F. Let us now estimate the error we made
in neglecting the kinetic energy difference in calculating the temperature rise.
The ratio of the final to initial specific volume is
v2 /v1 = ( p1 / p2 )(T2 /T1 ) = (20/16)(585.3/529.7) = 1.381
and then from the steady flow condition the final velocity is
v2 = (v2 /v1 )v1 = 4.143 ft/s
Accordingly the change in kinetic energy is 4.08 ft2 /s2 , which converts to
0.000163 Btu/lbm, and with the c p of air represents a 0.000676 ◦ F temperature
rise. This is certainly negligible compared with our calculated 55 ◦ F.
162 4 The First Law
Mixing Chambers and Heat Exchangers

A mixing chamber is a device in which two inlet streams of fluid are combined, and
the resulting mixture exits in a steady flow. There is no work done in this three-state
process, and, unless stated to the contrary, there is no heat transfer to the surroundings.
The following example illustrates this type of problem.
Example 4.13 A stream of cold water at 40 ◦ F, 20 psi is mixed with a hot

stream at 160 ◦ F, 40 psi. If the outlet stream is to be at 100 ◦ F, 14.7 psi, what
is the ratio of the mass flow rates of the hot stream to the cold?
Solution.
State 1 State 3
p1 = 40 psi p3 = 14.7 psi
Tf 1 = 160◦ F Ṁ3 h3 =
Ṁ1 h1 +
State 2 Ṁ2 h2 +
Ṁ2 = p2 = 20 psi Q̇ − Ẇ
Ṁ3 − Ṁ1 Tf 2 = 40◦ F Tf 3 = 100◦ F (energy
(steady flow) equation)
Ẇ = 0 (no moving surface)

Q̇ = 0 (adiabatic)
Ten of the 11 equations required to completely specify this three-state problem

are written in terms of the given information. Therefore, the problem cannot
be solved completely, however as we have noted before, we can solve for the
intensive variables in this situation, for example, the ratio of the mass flow
rates of the hot stream to the cold, r = Ṁ1 / Ṁ2 can be found by substituting
mass conservation into energy conservation and using q = w = 0
( Ṁ1 + Ṁ2 )h 3 = Ṁ1 h 1 + Ṁ2 h 2
dividing by Ṁ2 , and solving for r . The result is
r = (h 2 − h 3 )/(h 3 − h 1 )
From the given temperature, the enthalpy in the three states are found from the
tables as 128.07, 8.086, and 68.04 Btu/lbm, respectively. These values used in
the equation
r = (8.086 − 68.04)/(68.04 − 128.07)
produce the value 1.00; equal flow rates of 160◦ water and 40◦ water produce
100◦ water.
A heat exchanger is a four-state device in which a hot fluid is cooled by a cold

fluid that is simultaneously heated. It is the open system equivalent of a calorimetric
process (refer to Sect. 3.3.3). There is no work done in this process and normally the
heat transfer from the hot fluid is equal to the heat transfer to the cold. That is the
heat exchanger is thermally isolated. In those problems for which we want to know
the rate of heat transfer, we must analyze each fluid separately.
Example 4.14 In an automobile radiator, hot water flowing in tubes at a rate of

80 l/min is cooled from 100 ◦ C, 120 kPa to 90 ◦ C, 100 kPa by passing flowing
cold air around the tubes. If the air enters at 40 ◦ C, 140 kPa, exits at 101 kpa,
and is flowing at a rate of 80 kg/min, what is the final air temperature, and
what is the rate of heat transfer to the air? At these low temperatures, air can
be considered a perfect gas.
Solution.
State 1A State 2A
p1A = 140 kPa p2A = 101 kPa Ṁ2A = Ṁ1A
Ṁ1A = (steady flow)
80 kg/min Tc1A = 40◦ C Q̇W + Q̇A
=0
(adiabatic)
ẆA = 0 (no moving surface)

Q̇A = Ṁ1A (h2A − h1A ) (energy-air)
State 1W State 2W
p1W = 120 kPa p2W =
V̇1W = 100 kPa Ṁ1W = Ṁ2W
.08 m3 /min (steady flow-
Tc1W = 100◦ C Tc2W = 90◦ F water)
ẆW = 0 (no moving surface)

Q̇W = Ṁ1W (h2W − h1W ) (energy-water)
We calculate the rate of heat transfer to the water using the enthalpy of water
at the two temperatures, 419.06 kJ/kg, and 376.95 kJ/kg as found in the steam
tables
⎧
⎨ Ṁ1W (h 2W − h 1W ) = (V̇1W /v1 )(h 2W − h 1W )
Q̇ W = (0.08 m3 /min)/(0.001036 m3 /kg)[(376.95 − 419.06)kJ/kg]
⎩
−3251.7 kJ/min
which is also the rate of heat transfer from the air. The final air temperature is
then determined by expressing it from the equation for Q̇ A using the adiabatic
condition and the perfect gas enthalpic equation of state for air
164 4 The First Law
⎧
⎨ Tc1A − Q̇ W /(c p Ṁ1A )
Tc2 A = 40 ◦ C − (−3251.7 kJ/min)/[(1kJ/kg ◦ C)(80 kg/min)]
⎩
80.6 ◦ C
The water enthalpy difference can be accurately approximated by the product
of the temperature difference and c p ∼ 4.19 kJ/kg K [the v O ( p2 − p1 ), see
Eq. (4.29), term is very small]. This is 41.9 kJ/kg instead of the 42.11 KJ/kg
that we used.
If the internal heat transfer rate is not required the problem is simpler, we then
consider the exchanger itself (both substances) as the system. The following is an
example of this type of problem.
Example 4.15 In an automobile radiator hot water flowing in tubes at a rate of

80 l/min is cooled from 100 ◦ C, 120 kPa to 90 ◦ C, 100 kPa by passing flowing
cold air around the tubes. If the air enters at 40 ◦ C, 140 kPa, exits at 101 kpa,
and is flowing at a rate of 80 kg/min, what is the final air temperature? At these
low temperatures air can be considered a perfect gas.
Solution.
State 1 State 2
p1 = 140 kPa p2 = 101 kPa
Ṁ1 = Ṁ2 h2 = Ṁ1 h1 + Ṁ2 = Ṁ1
80 kg/min Tc1 = 40◦ C Ṁ3 h3 − Ṁ4 h4 (steady air
(energy equation) flow)
State 3 State 4
p3 = 120 kPa p4 =
V̇3 = 100 kPa Ṁ3 = Ṁ4
80 l/min (steady
Tc3 = 100◦ C Tc4 = 90◦ C water flow)
w=0 (no moving surface)

q=0 (adiabatic)
Here we use the two mass conservation equations in the four state energy
equation, the perfect gas enthalpic equation of state, the approximation
c pW (Tc3 − Tc4 ) for the water enthalpy difference (see previous example), and
solve for the final air temperature
Tc2 = Tc1 + [V̇3 /(v3 Ṁ1 )](c pW /c p A )(Tc3 − Tc4 )
In terms of c pW = 4.19 kJ/kg K, c p A = 1 kJ/kg K, and v3 = 0.001036 m3 /kg,

this is
Tc2 = [40 + (.08)/(0.001036 · 80)(4.19/1)10] ◦ C
or 80.4 ◦ C in good agreement with what we got before.

Duct Flow
A nozzle is a device which converts the internal energy of a fluid to kinetic energy
by means of a varying flow area. High-pressure, low-speed fluid is taken in at the
inlet of large area, and low pressure, high speed fluid is exhausted at a section with
a small area. No work is done because no surface moves, and ideally the process is
adiabatic. Conversely, a diffuser is a device in which kinetic energy is converted to
internal energy. Nozzles and diffusers are examples of duct flow in which the flow
velocity along with the other flow properties change due to changes in the area of
the duct along its length. As you might imagine, in duct flow problems we usually
cannot neglect the kinetic energy terms in Eq. (4.63).
Example 4.16 The nozzle of a hose inlets water at 60 ◦ F, 20 psi, and the
volume rate of flow is 0.01 ft3 /s over an area of 0.196 in2 . At the exit, the water
temperature is the same as it was at the inlet, the pressure is atmospheric (14.7
psi), and the height is essentially the same as the inlet. Assuming that the flow
is adiabatic, what is the final velocity?
Solution.
State 1 State 2
p1 = 20 psi p2 =
V̇1 = 14.7 psi V̇2 = V̇1
0.01 ft3 /s Tf 1 = 60◦ F T2 = T1 (steady
(stated) flow)
A1 = 0.141 in2 v2 = [2(h1 − h2 + q − w) + v12 ]1/2
z1 = z2 (energy equation)
(stated)

q=0 (adiabatic)
Since we have written 11 equations here, there is enough information to deter-

mine all the variables, although we are not asked to do this. Since the temper-
atures are the same in the two states and water is incompressible Eq. (4.29)
gives
h 2 − h 1 = c p (T2 − T1 ) + v O ( p2 − p1 ) = v O ( p2 − p1 )
and the final velocity is then with q = w = 0
v2 = [2v O ( p1 − p2 ) + v12 ]1/2
The initial velocity is

⎧
⎨ V̇1 /A1
v1 = (0.01 ft3 /s · 144 in2 /ft2 )/(0.196 in2 )
⎩
7.33 ft/s
166 4 The First Law
The final velocity is, therefore, with v O = v f (60 ◦ F) = 0.16033 ft3 /lbm,
⎧
⎨ {2(0.16033 ft3 /lbm)[(20 − 14.7)lbf/in2 ](144 in2 /ft2 )
v2 = ·(32.17 lbm ft/lbf s2 ) + 7.3342 ft2 /s2 }1/2
⎩
29 ft/s
As I remarked before we could also calculate the final area, etc. if required.
Example 4.17 Steam enters a circular nozzle with pressure, temperature, and
flow speed of 80 psi, 320 ◦ F, and 3 ft/s respectively. At the exit the pressure is
20 psi, the quality is 95%, and the mass rate of flow is 1 lbm/s. Assuming the
flow is adiabatic, what is the final velocity?
Solution.
State 1 State 2
p1 = 80 psi p2 = 20 psi
Ṁ1 = Ṁ2 Tf 1 = 320◦ F Ṁ2 = 1 lbm/s
(steady x2 = 0.95
flow) v2 = [2(h1 − h2 + q − w) + v12 ]1/2
v1 = 3 ft/s energy equation

q=0 (adiabatic)
The initial enthalpy can be read from the table; it is 1187.5 Btu/lbm. The final
enthalpy can be calculated from the saturation properties, and given quality

196.27 + 0.95 · (1156.3 − 196.27)
h2 =
1108.3 Btu/lbm
The outlet velocity is therefore with q = w = 0

⎧
⎨ {2 · [(1187.5 − 1108.3) Btu/lbm] · (778.2 ft lbf/Btu)
v2 = ·(32.17 lbm ft/s2 lbf) + 9 ft2 /s2 }1/2
⎩
1996 ft/s
Note the use of the conversion factors here and in the previous problem to
obtain ft/s.
A throttling process is one in which a constriction is placed in a pipe of flowing

fluid. A throttle is used to drop the pressure in the fluid, a valve is an example of
a throttle. At a point downstream of the constriction where the state of the flowing
fluid can again be regarded as uniform across the cross-section, the kinetic energy is
nearly equal to its value upstream of the constriction, and there is little change in the
potential energy as well. Moreover the fluid passes through the throttle so quickly that
no significant amount of heat transfer occurs. Thus a throttle is a special case of duct
flow in which we can neglect the kinetic energy terms, and consequently we have
h2 = h1 (4.70)
This is the form of the energy equation that is used in throttling problems.
Example 4.18 Steam is throttled from 160 psi to 240 ◦ F, 14.7 psi. What was
the initial quality?
Solution.
State 1 State 2
p1 = 160 psi p2 = 14.7 psi
h1 = h2 Ṁ2 = Ṁ1
(energy Tf 2 = 240◦ F (steady
equation) flow)

q=0 (adiabatic)
The final enthalpy is found in the table of steam properties; it is h 2 = 1164

Btu/lbm. Therefore the initial state is specified by this enthalpy at a pressure of
160 psi. According to the tables, the water in these conditions is a liquid–vapor
equilibrium mixture at a temperature of 363.55 ◦ F. The data in the table for
the saturation curve give
x1 = (1164 − 336.07)/(1195.1 − 336.07) = 0.96
As you see here, a throttle can be used to obtain the quality of a (mostly vapor)
liquid–vapor mixture by making only temperature and pressure measurements.
This device is called a throttling calorimeter.
4.4.3 Variable Mass Systems
Another application for which we can obtain a relatively simple energy equation
from Eqs. (4.58) and (4.59) is filling a tank or flexible vessel from a reservoir. A
typical arrangement, shown in Fig. 4.5, is modeled by (neglecting any change in
location of the system mass center)
d MT /dt = Ṁ
dUT /dt = h R Ṁ + Q̇ T − ẆT
168 4 The First Law
Fig. 4.5 An illustration of the tank filling process
in which the properties of the reservoir that fills the vessel are constant throughout the
process (that is what we mean by a reservoir), and we have neglected the kinetic and
potential energy terms in Eq. (4.59). Eliminating Ṁ between these two equations,
integrating the result from an initial time, when the system is in its initial equilibrium
state, to the final time, when the system reaches its final equilibrium state, and noting
that h R is a constant throughout, we obtain
M2 u 2 − M1 u 1 = h R (M2 − M1 ) + Q T − WT (4.71)
This is the appropriate form of the energy equation to use in solving open system
problems of this type. The following is a general example.
Example 4.19 A rigid tank of volume V is filled with gas, from a line in which
the pressure is p ∗ and temperature T ∗ . The initial pressure is p < p ∗ and tem-
perature T . The process is adiabatic and continues until the pressures are equal.
Assuming the gas is perfect, derive an expression for the final temperature in
the tank.
Solution.
State 1 State 2
V1 = V p1 = p p2 = p∗ V2 = V1
(equilibrium) (rigid tank)
T1 = T M2 u2 − M1 u1 =
h∗ (M2 − M1 )
(energy equation)
WT = 0 (no moving surface)

QT = 0 (adiabatic)
The initial and final masses are obtained using the equation of state
4.5 Exercises 169
M1 T = pV /R M2 T2 = p ∗ V /R
Substituting into the energy equation and canceling produces
p ∗ − p = kT ∗ ( p ∗ /T2 − p/T )
and solving for T2 gives the expression
T2 /(kT ∗ ) = ( p ∗ / p)/( p ∗ / p − 1 + kT ∗ /T )
Note that the final gas temperature is always higher than T ∗ .
4.5 Exercises
In each of the following exercises where appropriate, indicate the change of state on
a suitable projection plane of the state surface. For substances that can undergo a
phase change sketch the saturation curve as well in this diagram.
Section 4.2.2
4.1 A vat of molasses, V , is stirred by a 2 hp motor and heated by a 50000 Btu/hr

burner over its bottom surface. If the system is in a steady state, how much heat
transfer is there from the molasses to the room, R? If the room is 70 ◦ F, the
vat is a 6 ft high 6 ft diameter cylinder, and the heat transfer coefficient h V /R is
3 Btu/(hr ft2 R) for the side and top of the vat, what is the temperature of the molasses
[use Eq. (3.47)]?
4.2 The shock absorber, S, of example 3.1 operates in the steady state, but at a
frequency of 10 Hz. What is the average rate of heat transfer to the surroundings, S̃?
If the shock is a 10 in long cylinder of diameter 12 in, the heat transfer coefficient
h S/ S̃ is 10 Btu/(hr ft2 R), heat transfer occurs over the cylindrical surface only, and
the ambient temperature is 70 ◦ F, what is the average oil temperature in the shock?
(2420 Btu/hr, 162 ◦ F)
4.3 A power plant is heated by burning coal in its boiler at 12·106 Btu/hr, and heats
river water in its condenser at 8·106 Btu/hr. How much power does the plant produce?
4.4 A refrigerator is heated from its cold box at 3000 Btu/hr, and heats its surround-
ings at 5545 Btu/hr. How much power is required to operate this device in the steady
state? (1 hp)
4.5 A heat pump cools its surroundings at 6 kW and heats a house at 8 kW. How
much power is required to operate this device in the steady state?
170 4 The First Law
Section 4.3
4.6 Derive Eq. (4.22) from Eq. (4.21)
Section 4.3.1
4.7 Use Eq. (4.30) to show that in an isothermal change of state of a vaporizable
liquid the enthalpy change is h 2 − h 1 ∼ v f ( p2 − p1 ) where v f is the saturated liquid
specific volume at the common temperature.
4.8 One half gallon of water is contained in a vessel with a movable lid at 18 psi and
73 ◦ F. The water is heated, while the lid moves in order to keep the pressure inside
constant, until the liquid just begins to boil. What is the final water temperature, and
how much heat transfer is required?
4.9 Five milliliters of benzene at 25 ◦ C are placed in a nonrigid container and shaken
for 10 minutes. If the shaker motor produces 0.5 W, the device is 90% efficient, and
the process is adiabatic, what is the final liquid temperature? (58.9 ◦ C)
Section 4.3.2
4.10 Derive Eq. (4.35)
4.11 Plot the surface h( p, T ), 100 ◦ F ≤ T ≤ 700 ◦ F, and 10 psi ≤ p ≤ 50 psi,
using the tables for air as an ideal gas. Use any convenient software.
4.12 Using the energy equation, derive the expression
Q = M R(T2 − T1 )(n − k)/[(k − 1)(n − 1)]
for the heat transfer to a perfect gas (specific heat ratio k) in a polytropic process
with exponent n.
4.13 How much energy is required to vaporize 1 kg water at 80 ◦ C? (2308.8 kJ)
4.14 How much energy is required to vaporize 1 lbm of R-12 at 100 psi?
4.15 One and a half lbm of argon is heated slowly in a piston-cylinder arrangement
from 30 psi and 300 ◦ F. The piston, which is loaded by its own weight, rises until its
height increases by 30% and there it encounters stops. The heating continues until
the final temperature is 895 ◦ F. How much heat is transferred to the substance in this
process? (84.8 Btu)
4.16 One and a half lbm of argon is heated in a piston-cylinder arrangement from
30 psi and 300 ◦ F until the piston height increases by 30% and the temperature is
895 ◦ F. How much heat is transferred to the substance if this process is polytropic?
30 psi and 300 ◦ F until the height increases by 30% and the temperature is 895 ◦ F.
If a linear spring is connected to the piston and the process takes place slowly, how
much heat is transferred to the substance? If the piston diameter is 8 in, what is the
value of the spring constant? (88 Btu, 5.32 lbf/in)
4.5 Exercises 171
30 psi and 300 ◦ F until the temperature increases by 300 ◦ F. If a linear spring with
constant 5.32 lbf/in is connected to the 8 in diameter piston and the process takes
place slowly, how much heat is transferred to the substance?
4.19 One and a half lbm of steam is heated in a piston-cylinder arrangement from
30 psi and 300 ◦ F until the temperature increases by 300 ◦ F. If a linear spring with
constant 5.32 lbf/in is connected to the 8 in diameter piston and the process takes
place slowly, what is the final pressure, and how much heat is transferred to the
substance? (36.9 psi, 165 Btu)
4.20 Steam is contained in a 1.7 ft3 tank at 600 psi and 500 ◦ F. A valve is slowly
opened and the steam expands into a larger volume. If the final pressure is 120 psi
and the final temperature is 374 ◦ F, how much heat is transferred to the steam?
4.21 Steam is contained in a 1.7 ft3 tank at 600 psi and 500 ◦ F. A valve is slowly
opened and the steam expands into a larger volume. If the final pressure is 120 psi
and the process is adiabatic, what is the final temperature? (387 ◦ F)
4.22 Saturated vapor steam is contained in a 1.7 ft3 rigid tank at 180 psi. If the steam
is slowly heated until the final temperature is 800 ◦ F, what is the final pressure, and
how much heat is transferred to the steam?
4.23 Two lbm of saturated liquid R-12 contained in a 6 in diameter piston cylinder
arrangement fitted with stops is heated from 20 psi until it exists as saturated vapor
at 120 ◦ F. How high does the piston rise before encountering the stops, how much
work is done by the R-12, and how much heat is transferred to it in the process? (2.29
ft, 1300 ft lbf, 152 Btu)
4.24 A rigid cylinder containing 1.5 kg of O2 at 125 kPa, 27 ◦ C, and another rigid
cylinder containing 2 kg of O2 at 147 kPa, 35 ◦ C are connected by a pipe with a
closed valve. If the valve is slowly opened, and the contents of the two cylinders mix
adiabatically, what is the final temperature and pressure?
Section 4.3.3
4.25 Two lbm of saturated water vapor in a rigid vessel is cooled until its pressure
is 20 psi, and quality 25%. What is the initial temperature, and how much heat is
transferred from the water in this process?
4.26 One and a third lbm water at 20% quality and 80 ◦ F in a closed nonrigid vat
is stirred until it exists as saturated vapor at 100 ◦ F. If this adiabatic process is 90%
efficient, and the motor produces 1 hp, how long did the motor run? (27.8 min)
4.27 A cylinder fitted with a pinned piston contains 30 l of R-12 at 60 ◦ C, 90%
quality. When the pin is removed the piston moves vertically to a new equilibrium
with the R-12 quality 87% and a pressure of 500 kPa. If the process is adiabatic, how
much work is done by the R-12? If there is a constant load on the piston, what is the
mechanical efficiency of the process?
172 4 The First Law
4.28 Three kg of water at 60 ◦ C and 1% quality is heated in a rigid tank until the
temperature is 200 ◦ C. How much heat is transferred to the water in this process?
(4910 kJ)
Section 4.4.2
4.29 Show that in a throttling process of an ideal gas the final temperature is equal
to the initial temperature.
4.30 Steam enters an adiabatic turbine at 1250 psi, 800 ◦ F and leaves at 5 psi, 90%
quality. What mass flow rate is required for a power output of 1 MW? (2.74 lbm/s)
4.31 Air flows steadily through an adiabatic turbine, entering at 150 psi, 900 ◦ F, and
300 ft/s, and leaving at 20 psi, 300 ◦ F, and 700 ft/s. If the inlet area of the turbine is
0.1 ft2 , what is the outlet area, mass rate of flow, and power output?
4.32 Water at 50 ◦ F and 60 psi is heated in a chamber by mixing it with saturated

vapor at 60 psi. If both streams enter the chamber with the same mass flow rate and
the final pressure is 60 psi, what is the temperature and quality of the exiting stream?
(293 ◦ F, 37.2%)
4.33 A feedwater heater in a steam power plant mixes steam at 140 psi, 360 ◦ F with
water at 140 psi, 100 ◦ F. What is the ratio of the mass flow rate of steam to water in
order that the exit mixture is saturated liquid at 140 psi?
4.34 A heat exchanger in an air conditioning system cools a stream of R-12 that
is initially at 200 psi, 100 ◦ F. The cooling stream of R-12 enters the exchanger as
saturated vapor at −20 ◦ F, and exits, at the same pressure, at 80 ◦ F. If the mass rate
of flow of the two streams is the same, and the hot stream exit pressure is the same
as the inlet, what is its exit enthalpy and temperature? (17.45 Btu/lbm, 39 ◦ F)
4.35 Air enters a nozzle at 50 psi, 140 ◦ F, 150 ft/s, and leaves at 14.7 psi, 900 ft/s.
The heat loss from the nozzle is 6.5 Btu/lbm of air flowing, and the inlet area is
0.1 ft2 . What is the exit temperature, and exit area of the nozzle?
4.36 Steam at 14.7 psi and 320 ◦ F enters a diffuser with a velocity of 500 ft/s, and
leaves as saturated vapor at 240 ◦ F with a velocity of 100 ft/s. If the exit area of the
diffuser is 120 in2 , what is the mass flow rate, the diffuser inlet area, and the rate of
heat transfer from the steam? (5.1 lbm/s, 46.1 in2 , 236 Btu/s)
4.37 R-12 at 120 psi, 100 ◦ F is throttled to 80 ◦ F. What is the final pressure and
internal energy?
4.38 Steam enters a turbine at 1 MPa, 550 ◦ C with a velocity of 60 m/s, and leaves
at 50 kPa, and a quality of 95%. If there is a heat loss of 30 kJ/kg during the process,
the inlet area is 50 cm2 , and the exit area is 820 cm2 , what is the mass flow rate, the
exit velocity, and the power output? (7.97 kg/s, 29.9 m/s, 815 kW)
4.5 Exercises 173
4.39 Air enters a 1200 w hairdryer at 100 kPa, 22 ◦ C, and leaves at 47 ◦ C, 100 kPa.
The cross-sectional area of the inlet is 25 cm2 , and is the same as the outlet; moreover,
the flow is adiabatic. Assume the air is a perfect gas and determine the volume rate
of flow at the inlet, and the exit velocity.
4.40 The free surface of the water in a well is 20 m below ground level, and this
water is to be pumped to an elevation of 30 m above ground level. The area of the
inlet and outlet pipes are the same, as are the two temperatures and pressures. If this
is an adiabatic process and the groundwater is at 20 ◦ C, what power is required to
drive the pump in order to deliver a volume flow rate of 1.5 m3 /min? (12.2 kW)
4.41 Air enters the evaporator section of an air conditioner at 100 kPa, 90 ◦ C and a
volume rate of flow of 12 m3 /min. R-12 at −20 ◦ C with a quality of 30% enters the
evaporator at a rate of 4 kg/min and leaves as saturated vapor at the same pressure.
The final air pressure is 100 kPa, what is the exit air temperature, and the rate of heat
transfer from the air to the R-12?
4.42 A hot water stream at 70 ◦ C enters a mixing chamber with a mass flow rate of
0.6 kg/s where it is mixed with a stream of cold water at 20 ◦ C. If it is desired that
the mixture leave the chamber at 42 ◦ C, what is the required mass rate of flow of the
cold water stream? Assume that all pressures are 100 kPa. (0.764 kg/s)
4.43 R-12 at 800 kPa and 80 ◦ C enters an adiabatic nozzle with a velocity of 15 m/s,
and leaves at 300 kPa and 22 ◦ C. What is the exit velocity, and the ratio of initial to
final area?
4.44 R-12 enters a diffuser as saturated vapor, 30 ◦ C, 120 m/s, and leaves at 800 kPa,
40 ◦ C. The refrigerant gains heat at a rate of 12 kW as it passes through the diffuser,
and the exit area is 30% greater than the inlet. What is the mass rate of flow, and the
exit velocity? (3.83 kg/s, 89.5 m/s)
Section 4.4.3
4.45 A 10 ft3 rigid tank contains water, at 2% quality and 450 ◦ F. If liquid is with-
drawn while maintaining the temperature constant, how much heat transfer is required
to halve the mass of water? (1700 Btu)
4.46 In the previous problem, how much heat transfer is required if vapor is with-
drawn instead of liquid?
4.47 An evacuated, insulated 15 m3 rigid tank is connected through a valve to a
steam line at 4 MPa, 300 ◦ C. If 45 kg of steam is admitted to the tank, and the final
temperature is 400 ◦ C, what is the final pressure and how much heat is transferred to
the steam? (0.93 MPa, −117 kJ)
4.48 A 200 ft3 rigid tank contains saturated vapor steam at 20 psi. Steam flowing
in a line at 100 psi, 400 ◦ F is admitted to the tank through a valve until the tank is
at the line pressure. What is the mass of steam that enters the tank in this adiabatic
process? What is the final temperature?
Chapter 5
The Second Law
5.1 Introduction
In our study till now, we established the first law of thermodynamics, in which we
saw that heat and work are equivalent insofar as they are different manifestations
of the energy exchange between a system and its surroundings (or between various
parts of a system). However, in this development, an asymmetry between heat and
work has remained. Recall from chapter 3 that in an equilibrium process, the rate of
doing work, or power developed, is given in Eq. (3.31)
dv p
Ẇ = M p p ,
dt
and the rate of heat absorption is given by Eq. (3.49)
dv p dT p
Q̇ = Mλv + Mcv .
dt dt
If we had chosen to represent the process in terms of the variables p, T in place of
v, T , we would have instead
dp p dT p
Ẇ = −M p p v p β + M ppvpα
dt dt
dp p dT p
Q̇ = Mλ p + Mc p .
dt dt
From this, we see that for a general choice of independent variables the rates of both
doing work and absorbing heat are linear functions of the rates of each variable, but
when one of the variables is specific volume, v, the power developed depends on
its rate only. There is no corresponding variable for which the rate of heat absorp-

176 5 The Second Law
tion is expressed in terms of its rate only; this is the asymmetry mentioned above.
As we shall see, removing this asymmetry has very far reaching consequences for
thermodynamics.
5.2 Entropy
In attempting to set up an expression analogous to Eq. (3.31) for the rate of heat
absorption, it seems natural from what we learned about the relationship between
pressure and temperature to mechanical and thermal equilibrium in Chaps. 1, 2,
and 3, to use absolute temperature in place of pressure. Moreover, although there
is no physically meaningful variable that can replace volume, the label s that was
used in Sect. 3.3.5 to describe an adiabatic equilibrium process, can be used for this
purpose.
Situations like this often occur in science, when either a mathematical develop-
ment or a sense of theoretical simplicity seems to demand the existence of something
that has not been discovered by experiment. In such cases, introduction of the appro-
priate theoretical construct can reorganize known facts in new ways that enhance
understanding, and integrate concepts that were previously thought to be unrelated1 .
In the present instance, the asymmetry we have noted suggests that a new variable
should exist, and one has been proposed. Here we introduce it by means of an Axiom.
As we shall see, this step has the expected beneficial consequences.
Axiom 5.1 (Existence of Entropy) For all closed thermodynamic systems that
admit macroscopic equilibrium states there exists an extensive property, S, called
entropy such that
1. in an equilibrium state, the system entropy is, apart from a constant, a function
of state and,
2. during an arbitrary equilibrium process, the rate of increase of the entropy is
equal to the rate of heat absorption by the system divided by the temperature
dS Q̇
= (5.1)
dt T
Credit for being the first to introduce entropy is usually given to Rudolf Clausius,
although W. M. Rankine also developed the concept at about the same time. Accord-
ing to Axiom 5.1, entropy has dimension [E/T] = [ML2 /Tt2 ] and, like energy, is a
relative quantity.
1 Newton’s law of Universal Gravitation, Eq. (1.25), is a prime example. The introduction of a
“force of gravity”, that acted over enormous reaches of empty space, together with the law of
motion, Eq. (1.26), not only gave a quantitative description of celestial motion, but also showed
how the planetary orbits were related to the motion of projectiles in the vicinity of the earth.
5.2 Entropy 177
The first part of Axiom 5.1 specifies S as an extensive state function. This means
that for a simple compressible substance in a macroscopic equilibrium state there is
an entropic equation of state of the form
S = S(v, T, M) = Ms(v, T ) (5.2)
where s, the specific entropy, is an intensive property. In the following section, we

will derive expressions for this function in terms of already familiar properties.
Once the entropy has been tabulated, the way we did for internal energy, it can
be used as we have the other thermodynamic variables in order to solve practical
problems. The most important applications of this type are the following:
• In an equilibrium isothermal process, Eq. (5.1) integrates to
Q = T (S2 − S1 ) (5.3)
Therefore, two equilibrium isothermal processes of a system, S, that take place at

temperatures TH and TL , and for which the entropy difference is the same, satisfy
the relation
TH Q S/H
= (5.4)
TL Q S/L
where Q S/H and Q S/L are the heat transferred to S at the high and low temper-
atures, respectively. Accordingly, the absolute temperature ratio is given by the
measured heat absorption ratio, independent of the substance in the thermometer,
the substance in the calorimeter, or the substance that undergoes the processes. In
other words, the introduction of entropy provides us with a definition of temper-
ature that is freed of its connection to ideal gases, or to any other substance. For
this reason, we call it the thermodynamic temperature.
• An isentropic process is one which is both equilibrium and adiabatic. As we
shall see, such processes, in which the entropy is constant [see Eq. (5.1) and
Sect. 3.3.5], are worth considering because they are neither mechanically, nor
thermally irreversible. Therefore, they can serve as idealizations against which to
compare the efficiency of our own designs.
Entropy can also be used to advantage as a variable for plotting processes. Just as the
work done in a quasi-equilibrium process is the area under the process curve on a p, v
plane, the heat absorbed in a quasi-equilibrium process is the area under the process
curve in a T, s plane. This is shown in Fig. 5.1. Moreover, for a quasi-equilibrium
cyclic process, the area enclosed by the process curve in this plane is the net work
done in the process, as required by Corollary 4.3. In order to make use of entropy
as a thermodynamic property for a specific substance, we need to know its entropic
equation of state.
Fig. 5.1 An illustration of the heat transfer in a quasi-equilibrium process as the area under the
curve in the T, s plane
Kinetic Model for Entropy

The simple kinetic model we used for pressure, N p particles of mass m p , enclosed in
a cubical volume V , with one-third of the particles traveling in each direction with
the same speed (total energy U = N p m p c2 /2), and not colliding with one another, is
too simple to give a kinetic account of entropy. Indeed the role of collisions between
particles is important and cannot be neglected, because the result of a collision
between two particles having the same speed is that each particle emerges with a
speed different than the other and of their common original speed. After a large
number of collisions, there will be particles having squared speeds from zero to
N p c2 (consistent with the requirement that the total energy remain U ). We describe
this situation by thinking about the space of the coordinates and velocity components
of a particle at a given moment (x, y, z, cx , c y , cz ) and divide this space into J cubes
all of the same volume. Then, at any instant, each particle will be located in one of
these cubes, which we can number 1, 2, · · · j · · · J and the entire collection can then
be described by the number of particles in each cube N1 , N2 , · · · N j , · · · N J . Each
possible distribution is constrained by the requirements that

J
1 1 J
Np = Nj U= m p N p c2 = m p N j c2j
j=1
2 2 j=i
namely, that the total number of particles is N p and the total energy is U (since
collisions do not change either the number of particles or their energy). The quantity
c2 is the mean square speed. It turns out that the relations we obtained for p and T in
Eqs. (1.41) and (1.50) also remain true if we replace c2 in them with c2 . Now all the
cubes have the same volume, so if we assume that the particles are randomly located
with random velocity, the probability of a particle being in any volume j is the same
for all of them. In this circumstance, Ludwig Boltzmann noted that the probability of
realizing a particular distribution of identical particles is proportional to a quantity,
Z , that denotes the number of ways of realizing it, and given by
5.2 Entropy 179

J
1
Z = N p!
j
N j!
In this equation, the symbol N p ! denotes the factorial function, the product of all
the integers from 1 to N p . The expression on the right, known from the theory of
permutations and combinations, produces the number of distinct permutations of N p
objects taken N p at a time, when N1 of them are identical (in cube 1), N2 are identical
(in cube 2), etc. Boltzmann is also credited with proposing the equation
S = k B ln Z + constant (5.5)
According to Eq. (5.5), the most probable distribution has the largest entropy and,
as such, entropy is simply a measure of the probability of a particular particle distri-
bution. From this point of view Clausius’ dictum, “. . . the entropy of the Universe
tends toward a maximum”, is simply a statement of a law of probability. Note that
the distribution in which all particles are located in one cell has, apart from a con-
stant, S= 0 (in this case Z = 1, moreover the cell j = j ∗ must be the one for which
c j ∗ = c2 due to the condition on U ). Since this represents the highest possible state
of order, whereas the most probable distribution (that has the largest value of S) can
be thought of as least orderly, entropy is also considered a measure of the degree of
disorder of the particle distribution.
This theory, which was pioneered by Boltzmann, Maxwell, and Gibbs, around the end
of the nineteenth and beginning of the twentieth centuries, has been developed into
the subject of Statistical Thermodynamics, in which the specific heats and equations
of state of substances are developed based on their atomic and molecular structure.
5.3 The Entropic Equation of State
Here, we use an equilibrium process specified by the independent variables v and T ,
v = v p (t) T = T p (t) p = p p (t) = p[v p (t), T p (t)]
in order to construct s = S/M = s p (t) (we did a similar calculation to establish the
energetic equation of state in Sect. 4.3) by equating Eqs. (3.49) and (5.1) for the rate
of heat absorption by the system
ds p cv dT p λv dv p
= + (5.6)
dt T p dt T p dt
As we have seen before (Sects. 2.3.2, 2.4, 4.3), this is the differential equation of a
space curve that describes the process, and lies in the surface that defines the entropic
equation of state
s = s(v, T )
Consequently2
∂T s|v = cv /T ∂v s|T = λv /T
and
∂v (cv /T )|T = ∂T (λv /T )|v (5.7)
Carrying out the differentiations indicated in Eq. (5.7) produces
1 1 1
∂v cv |T = ∂T λv |v − 2 λv
T T T
and eliminating ∂v cv |T between this and Eq. (4.20), ∂v cv |T = ∂T λv |v − α/β, gives
λv = T α/β (5.8)
This extremely important result tells us that λv is determined by the mechanical

equation of state, and consequently all thermodynamic information about a simple
compressible substance is contained in only two functions, say p(v, T ) and cv (v, T ).
The specific heat at constant pressure, c p , is determined in terms of these two. We
previously saw that this was true for ideal gases (Sect. 4.3.2) based on Joule’s free
expansion experiments. However, here the result holds for all simple compressible
substances as a direct consequence of Axioms 4.1 and 5.1.
Equation (5.8) can be used to obtain many other useful relations. For example,
βλv = αT , so for liquids and solids βλv <<1 (see Sects. 2.4.1, Constant Property
Liquids and 4.3.1). From Eq. (4.18), ∂v u|T = λv − p, we get
∂v u|T = T α/β − p (5.9)
from Eq. (3.53), λ p = −vβλv ,

λ p = −vαT (5.10)
and from the expression on the previous page, ∂v s|T = λv /T ,
∂v s|T = α/β = ∂T p|v
This equation is called a Maxwell relation in honor of James Clerk Maxwell. It is

one of four so named relations that express a p or v partial derivative of the entropy
function in terms of measurable quantities.
2 Because of what will follow we need to specify here which variable is held constant in the partial
differentiation.
5.3 The Entropic Equation of State 181
Finally, from Eq. (3.52), c p = cv + vαλv ,
c p = cv + T vα2 /β (5.11)
This is perhaps the most striking equation in thermodynamics, insofar as it relates

six experimentally determinable quantities. Thus, it has been used to verify that
Axioms 4.1 and 5.1 are consistent with experience for many simple compressible
substances under a variety of conditions.
On substituting Eq. (5.8) into Eq. (5.6) and integrating, as we did before in chap-
ters 2 and 4 [see Eqs. (2.18) to (2.23)], we obtain an expression for the entropic
equation of state in terms of its experimentally determinable first partial derivatives
v
T
cv (v O , TC ) α(vC , T )
s = s(v, T ) = s O + dTC + dvC (5.12)
TO TC vO β(vC , T )
The arbitrary constant here is of no consequence because all physically meaningful

information is based on differences in entropy, for example, Eq. (5.3) for the heat
absorbed in an isothermal process.
Alternative Entropic Equation of State
It is often more convenient to use independent variables p and T to define s (refer to
Sects. 2.4 and 4.3). In that case, by using Eq. (3.50) in place of (3.49), and proceeding
in a completely analogous fashion, we obtain
ds p c p dT p λ p dp p
= +
dt T p dt T p dt
so that in this case the space curve lies in the surface
s = s[v( p, T ), T ] = ŝ( p, T )
Therefore, we have3
∂T s| p = c p /T ∂ p s|T = λ p /T
and
∂ p (c p /T )|T = ∂T (λ p /T )| p
so that finally

T
c p ( p O , TC ) p
s = ŝ( p, T ) = s O + dTC − v( pC , T )α( pC , T ) dpC (5.13)
TO TC pO
3 Note the difference between ∂ = c p /T and ∂T s|v = cv /T , which is why we must distinguish
T s| p
between them.
We obtain no new information from the equality of the mixed partial derivatives
in this case; on using it we simply recover Eq. (5.10). However, on substituting
Eq. (5.10), λ p = −vαT , into the expression for ∂ p s|T , we obtain
∂ p s|T = −vα = −∂T v| p
which is another of the four Maxwell relations.
5.3.1 Liquids and Solids
As we have noted before, liquids and solids are relatively incompressible. This means
that αT <<1 (see Sect. 2.4.1, Constant Property Liquids), and therefore a good
approximation to their entropic equation of state is obtained by ignoring the last
integral in Eq. (5.13). Furthermore over a limited range of T , c p is essentially constant
and we obtain simply from Eq. (5.13)
s ∼ s O + c p ln(T /TO )
In any event, we are interested only in differences of entropy values, and obtain
s2 − s1 ∼ c p ln(T2 /T1 ) (5.14)
independent of the values chosen for s O and TO . The following example shows how
we can use the entropy function in problems involving flowing liquids.
Example 5.1 A pump inlets a liquid at temperature T and pressure p at a

volume rate of flow V̇ over an area A. The pressure of the outlet stream
is p + p, the process is isentropic, the final area is A + A, the height
difference is zero, and steady flow conditions prevail. Derive an expression for
the power required to drive the pump.
Solution.
In writing the steady flow equations, we have made use of the fact that the
substance is an incompressible liquid, namely, v1 = v2 = v O . Moreover from
Eq. (5.14), the isentropic condition gives
0 ∼ c p ln(T2 /T1 )
namely, that the flow is approximately isothermal T1 ∼ T2 . In this case, we

find on using Eq. (4.29), h 2 − h 1 ∼ c p (T2 − T1 ) + v O ( p2 − p1 )
h 1 − h 2 ∼ v O [ p − ( p + p)] = −v O p
and therefore the power required by an external source is with Q̇ = 0
Ẇ M ∼ V̇ p + V̇ ρ O v22 /2[1 − (1 + A/A)2 ]
In many practical applications, the two areas are nearly equal so the second
term is ignorable. The work done per unit mass rate of flow is then simply
w M ∼ v O p
This formula is often used as an approximation in pump problems.
An important point to note here is that an isentropic process of an incompressible

substance is very nearly isothermal. This is the reason that thermal effects are ignor-
able for most problems you have studied (or will study) in the mechanics of solids,
and flowing liquids.
Vaporizable Liquids
As we have already seen in chapters 2 and 4, real liquids are both incompressible
and vaporizable. Therefore, we modify the path of integration in Eq. (5.13) to follow
the saturation curve from a selected reference point to the desired temperature, and
subsequently to follow an isothermal to the desired pressure (the path shown in
Fig. 2.7). Then Eq. (5.13) is
p
s = ŝ( p, T ) = s f − α( pC , T )v( pC , T ) dpC
ps (T )
where
ps (T )
T
c p [ ps (TC ), TC ]
s f (T ) = s O + dTC − α[ pC , Ts ( pC )]v̂ f ( pC ) dpC
TO TC pO
(5.15)
is the specific entropy of the boiling liquid. We make use of the incompressibility
by neglecting the integral in the equation of state, an approximation which is quite
accurate up to very large pressures. Accordingly, we obtain
s = ŝ( p, T ) ∼ s f (T ) p ≥ ps (T ) (5.16)
Thus the specific entropy, like specific volume, and specific internal energy depend
only on temperature, and is equal to the value of the entropy of the boiling liquid, s f
at that temperature. This information is tabulated along with the energy and enthalpy
as we discussed in chapter 4. In calculating specific values for the entropy using
Eqs. (5.15) and (5.16), we must make choices for s O , p O , and TO , but these arbitrary
decisions do not affect the value of the entropy difference between two states, as you
can see by doing the subtraction
ps (T2 )
T2
c p [ ps (TC ), TC ]
s f (T2 ) − s f (T1 ) ∼ dTC − α[ pC , Ts ( pC )]v̂ f ( pC ) dpC
T1 TC ps (T1 )
The following example shows how these tables can be used in the solution of prob-
lems.
Example 5.2 A boiling hot mug of coffee 4 in diameter and at atmospheric

pressure (14.7 psi) is allowed to cool at constant pressure to 120 ◦ F so that
it can be drunk. Initially the coffee stands 8 in high in the mug, there is no
loss of liquid by vaporization, and the thermodynamic properties of coffee are
the same as water. With these assumptions, calculate the heat transfer to the
surrounding air and the entropy change of the coffee.
As in all isobaric processes, the energy equation combined with the work
formula gives
Q = M1 (h 2 − h 1 )
a fact we have noted before (e.g., Example 4). The properties at state 1 are those
of boiling liquid at 14.7 psi. The tables give these as 0.01672 ft3 /lbm, 180.16
Btu/lbm, and 0.3121 Btu/lbm R for v1 , h 1 , and s1 , respectively. Moreover the
values of v2 , h 2 , and s2 at 14.7 psi and 120 ◦ F are 0.016204 ft3 /lbm, 88.01
Btu/lbm, and 0.1646 Btu/lbm R, respectively, which are also the values of the
saturated liquid at 120 ◦ F. The mass of liquid is given by V1 /v1
M1 = (32π in3 )/[(1728 in3 /ft3 )(0.01672 ft3 /lbm)] = 3.48 lbm
The heat transfer to the liquid is given by the energy equation
Q = (3.48 lbm)[(88.01 − 180.16)Btu/lbm] = −320.68 Btu
and the change in entropy of the liquid is
S2 − S1 = (3.48 lbm)[(0.1646 − 0.3121)Btu/lbm R] = −0.5149 Btu/R
Use of Eq. (5.14) here with c p = 1 Btu/lbm R as an approximation gives

−0.5126 Btu/R, which is within 5% of the more accurate calculation. We will
soon see what use we can make of this calculation of the entropy change.
5.3.2 Gases
The development of the entropic equation of state for gases follows along the lines
used in Sects. 2.4.2 for v and 4.3.2 for u and h.
Ideal Gases
The general form of the entropic equation of state, Eq. (5.12) can be somewhat
simplified for the special case of an ideal gas
v = RT / p
Indeed, as we saw in Eq. (2.42), this equation of state produces the quantities
α = 1/T β = 1/ p
which substituted into Eq. (5.9), ∂v u|T = T α/β − p, gives
∂v u|T = 0
This implies that u = u(T ), in agreement with the result of kinetic theory and Joule’s
experiments that we discussed in the previous chapter. This agreement follows as
a result of our choice of the ideal gas temperature as the absolute temperature in
Axiom 5.1. With these expressions substituted into Eq. (5.12), it takes the form
v
T
cv (TC ) dvC
s(v, T ) = s O + dTC + R .
TO TC vO vC
in which the last integral can be evaluated

T
cv (TC ) v
s = s(v, T ) = s O + dTC + R ln (5.17)
TO TC vO
Using Eq. (4.35), c p = cv + R, and the ideal gas equation of state, or alternatively,
evaluating the last integral in Eq. (5.13), this can be written as a function of the
variables p and T
(0) pO
s = ŝ( p, T ) = s (T ) + R ln (5.18)
p
where s (0) is given by

T
c p (TC )
s (0) (T ) = s O + dTC (5.19)
TO TC
The quantity s (0) is tabulated along with values for u and h in tables for specific ideal
gases. It is used by applying Eq. (5.18) between two equilibrium states
s2 − s1 = s2(0) − s1(0) + R ln( p1 / p2 ) (5.20)
in which case the arbitrary constant s O and reference values p O and TO used in
constructing the table are immaterial, as in the previous instances we have discussed.
For an isentropic process, Eq. (5.20) can be inverted to obtain
(0)
(T )−s (0) (TO )]/R T −[s (0) (T )−s (0) (TO )]/R
pr (T ) = pr (TO )e[s or vr (T ) = vr (TO ) e
T0
The functions pr and vr are calculated by choosing pr (TO ) = vr (TO ) = 1 and tab-
ulated along with u, h, and s (0) . These functions are used only for solving problems
involving isentropic processes by noting from their definition that
p2 / p1 = pr 2 / pr 1 or v2 /v1 = vr 2 /vr 1
Whether we use pr or vr depends on the problem statement. A common instance of

this procedure is illustrated in the following example.
Example 5.3 Air expands isentropically in a turbine from an initial pressure

of 100 psi and temperature of 700 ◦ F, to a pressure of 20 psi. Calculate the
final temperature, and the work done per pound of air. If the diameter of the
inlet feed pipe is 0.5 ft, the final area is 20% greater than the initial, and the
initial velocity is 20 ft/s, how much power is developed?
Solution.
From the table of properties of air, the initial h and pr (we use pr because p2
is given) at 700 ◦ F are
h 1 = 172.34 Btu/lbm pr 1 = 27.355 Btu/lbm R
Moreover with R = 53.34 ft lbf/lbm R (0.06854 Btu/lbm R) for air, the equa-
tion of state gives the value for v1

(53.34 ftlbf/lbmR)(1159.7 R)/[(144 in2 /ft 2 )(100 lbf/in 2 )]
v1 =
4.296 ft 3 /lbm
The mass rate of flow is then
Ṁ = (0.1963 ft2 · 20 ft/s)/(4.296 ft3 /lbm) = 0.9139 lbm/s
and pr 2 = 27.355 · 20/100 = 5.471 Using this in the table gives us the final
temperature and enthalpy by interpolation

275 + [(300 − 275)/(5.8898 − 5.2299)](5.4710 − 5.2299)
Tf2 =
284.1 ◦ F (744.3 R)

66.534 + [(72.620 − 66.534)/(5.8898 − 5.2299)](5.4710 − 5.2299)
h2 =
68.758 Btu/lbm
The work done per lbm is then with Q̇ = 0
w = (172.34 − 68.76) Btu/lbm = 103.58 Btu/lbm
and the power is 94.7 Btu/s (134 hp).
Perfect Gases
For a perfect gas, the remaining integral in Eq. (5.17) can also be evaluated
s = s(v, T ) = s O + cv ln(T /TO ) + R ln(v/v O )

On using Eq. (4.37), cv = R/(k − 1), this can be written more compactly as
1/(k−1)
T v
s = s O + R ln (5.21)
TO vO
The alternative equation of state can be obtained similarly by evaluating Eq. (5.19)
and substituting into Eq. (5.18). It can also be obtained, without an integration, by
substituting the equation of state, v = RT / p, into Eq. (5.21)
s = ŝ( p, T ) = s O + R ln[(T /TO )k/(k−1) ( p O / p)] (5.22)
A third form, s[v, T ( p, v)] = s̃( p, v), of the entropic equation of state4 is obtained
here by substituting T = pv/R, into Eq. (5.12)
s = s̃( p, v) = s O + cv ln[( p/ p O )(v/v O )k ] (5.23)
Of course, any one of these expressions is used by subtracting values corresponding to

two equilibrium states, thereby eliminating the arbitrary constant, s O , and reference
values, p O , v O , TO . For example, Eq. (5.23) gives
s2 − s1 = cv ln[( p2 v2k )/( p1 v1k )]
Moreover when we use these equations in connection with an isentropic change in

state, s2 = s1 , they simplify still further,
k/(k−1) k/(k−1) 1/(k−1) 1/(k−1)
p2 v2k = p1 v1k p1 T2 = p2 T1 T2 v2 = T1 v1
The left and right equations here were obtained previously, when we discussed quasi-
equilibrium adiabatic processes (see Sect. 3.3.5).
An example of the use of these equations, and a comparison of the analysis of a
perfect gas with an ideal gas is shown in the following example.
Example 5.4 Air expands isentropically in a turbine from an initial pressure

final temperature, and the work done per pound of air. If the diameter of the
inlet feed pipe is 0.5 ft, the final area is 20% greater than the initial, the air is
treated as a perfect gas with specific heat ratio 1.4, and the initial velocity is
20 ft/s, how much power is developed?
form s = s̃( p, v) is especially useful in piston-cylinder problems. The last two (of the four)
4 The
Maxwell relations equate its partial derivatives, ∂ p s|v , and ∂v s| p to measurable quantities.
Solution.
For an isentropic process, we obtain from Eq. (5.22) (see above)
T2 = ( p2 / p1 )(k−1)/k T1
which evaluates to
T2 = (0.2)(0.4/1.4) (1160 R) = 732.4 R (T f 2 = 272.7 ◦ F)
and differs by 4% from the value obtained in the previous problem from the
tables of air properties. The initial specific volume is the same as in the previous
problem, 4.297 ft3 /lbm, therefore so is the mass rate of flow. The work done
per lbm of air is with Q̇ = 0
w = (0.241 Btu/lbmR)(1160 − 732.4) R = 103.0 Btu/lbm
which is less than 1 percent different from the value calculated using a variable
specific heat. The power developed is 94.1 Btu/s (133 hp).
Condensible Gases
Again our discussion parallels what we have already done in chapter 4. Since it is
convenient to use p and T as independent variables, we begin with Eq. (5.13), and
integrate at constant pressure from the saturated vapor curve to the state in question.
In this way, we obtain
T
c p ( p, TC )
s = ŝ( p, T ) = sg + dTC T ≥ Ts ( p)
Ts ( p) TC
Here sg is not arbitrary. It is related to s f through a vaporization process. Since

vaporization occurs isothermally, we can use Eq. (5.1) along with knowledge of
the amount of heat required to vaporize a unit mass of liquid, Eq. (4.41), Q/M =
h g − h f , to obtain
T (sg − s f ) = h g − h f (5.24)
Once s is tabulated, we can use it as we have the other properties to solve problems.
Example 5.5 Steam expands isentropically in a turbine from an initial pressure

final temperature, and the work done per pound of steam. If the diameter of
the inlet feed pipe is 0.5 ft, the final area is 20% greater than the initial, and
the initial velocity is 20 ft/s, how much power is developed?
Solution.
From the tabulated values, we find in state 1
h 1 = 1379.5 Btu/lbm v1 = 6.833 ft3 /lbm s1 = 1.8036 Btu/lbmR
The mass rate of flow is obtained from the specific volume and is smaller than
that of air because steam has a smaller molecular weight (smaller density)
Ṁ = (0.1963 ft2 · 20 ft/s)/(6.833 ft3 /lbm = 0.5746 lbm/s
State 2 is located by interpolation using the chord line for 20 psi that brackets
the known value of entropy. In this way we find

320 + [(360 − 320)/(1.8170 − 1.7930)](1.8036 − 1.7930)
Tf2 =
337.7 ◦ F

22.976 + [(24.206 − 22.976)/(1.8170 − 1.7930)](1.8036 − 1.7930)
v2 =
23.34 ft3 /lbm

1201 + [(1220.1 − 1201)/(1.8170 − 1.7930)](1.8036 − 1.7930)
h2 =
1209.4 Btu/lbm
The work done per lbm is then with Q̇ = 0
w = (1379.5 − 1209.4) Btu/lbm = 170 Btu/lbm
This value is quite different from what we found previously for air due to the
difference in density of the two substances, however the power is 97.7 Btu/s
(138 hp), which compares closely with the value for air.
5.3.3 Liquid–Vapor Equilibrium

For an equilibrium of phases, an appropriate expression for the entropy is obtained by
means of the same procedure that we used previously (refer to Sects. 2.4.3 and 4.3.3)
for the other extensive variables. Thus from
S = S f + Sg = M f s f + Mg sg
we obtain the equation of state of the equilibrium combination on dividing by M.

This result is
s = (1 − x)s f + xsg = s f + x(sg − s f ) (5.25)
and is used in the same way as the corresponding expressions for v, u, and h,
Eqs. (2.47), (4.42), and (4.43) were previously.
Example 5.6 Steam expands isentropically in a turbine from an initial pressure

of 100 psi and temperature of 700 ◦ F, to a pressure of 5 psi. Calculate the final
temperature, and the work done per pound of steam. If the diameter of the inlet
feed pipe is 0.5 ft, the final area is 20% greater than the initial, and the initial
velocity is 20 ft/s, how much power is developed?
Solution.
The state 1 conditions are the same as in the previous problem

h 1 = 1379.5 Btu/lbm v1 = 6.833 ft3 /lbm s1 = 1.8036 Btu/lbmR
and the mass rate of flow is the same as well 0.5746 lbm/s. State 2 is a liquid–
vapor equilibrium because the entropy is less than the saturated vapor entropy
at 5 psi. Consequently, the final quality can be calculated from Eq. (5.25) using
data from the saturation curve
x2 = (1.8036 − 0.2349)/(1.8443 − 0.2349) = 0.9747
the final enthalpy then follows from this
h 2 = 130.20 + 0.9747 · 1000.9 = 1105.8 Btu/lbm
and the work done per lbm is, with Q̇ = 0
w = (1379.5 − 1105.8) Btu/lbm = 273.7 Btu/lbm
Here the work is larger than in the previous example, because of the additional
expansion down to 5 psi. Since the mass flow rate is the same as in the previous
example, the power developed here is 157 Btu/s (222 hp).
Equation (5.24) can be recast as

h f − T s f = h g − T sg
and in this form it is what we intended in writing Eq. (2.48). Thus, the existence of
entropy also provides the means for expressing the condition of material equilibrium.
The combination h − T s is given a symbol, g, and a name, the Gibbs free energy.
This is done because, like enthalpy, this combination of properties appears in a variety
of thermodynamic contexts.
5.4 The Irreversibility Principle
Until now, we have described processes by giving their initial and final states. In doing
this, we have observed a naturally occurring order imposed by our experience; for
example, when a hot body and a cold body, are brought into thermal contact, the hot
one is cooled and the cold one is heated until they both reach the same temperature,
and not vice versa.
It turns out that the entropy function, in addition to providing
• a basis for thermodynamic temperature,
• a measure for quasi-equilibrium, adiabatic processes,
• a state plane (together with absolute temperature) in which the area under a quasi-
equilibrium curve is the heat transfer,
5.4 The Irreversibility Principle 193
• a reduction in the number of measurable equilibrium properties required to specify

all thermodynamic information about a substance,
• a means for expressing phase equilibrium,
identifies this natural order and thus provides us with a timewise direction for the
progress of processes. This happens because contrary to the simplest possible exten-
sion of Axiom 5.1 to encompass nonequilibrium, adiabatic ( Q̇ = 0) processes
dS ?
= 0
dt
experience has shown that only one of the inequalities occurs in nature (note that
= means either > or <). This leads us to establish another axiom based on this
experience, however, in order to make it apply to a general nonequilibrium process,
for which no single value of T describes the system, we first need to introduce the
concept of entropy transfer.
5.4.1 Entropy Transfer
Whenever there is heat transfer to a system at a rate, −n j · q̇ j A j , over a part of its

bounding surface (see Sect. 3.3.1), there is concurrently an entropy transfer to the
system, at a rate
−n j · q̇ j A j /T j
Here T j is the local temperature, which is common to the two systems in contact.
The sum of all of these, in the limit, is a surface integral

−n · q̇
Ṡ =
T
dA (5.26)
A T
where we have denoted the rate of entropy transfer to the system by the symbol Ṡ T .
In terms of this quantity, we can formulate the second law of thermodynamics by
means of the following axiom.
Axiom 5.2 (Second Law) For all closed thermodynamic systems that admit macro-
scopic equilibrium states there exists an extensive property, S, called entropy such
that
1. in an equilibrium state, the system entropy is, apart from a constant, a function
of state and,
2. during an arbitrary process, the rate of increase of entropy is greater than or
equal to the rate of entropy transfer
dS
≥ Ṡ T (5.27)
dt
Historically, the irreversibility inherent in the inequality written in this axiom

was first proposed by Rudolf Clausius around 1850, who expressed it in terms of the
impossibility of heat transfer from any given temperature to a higher one occurring
spontaneously. William Thomson also developed the concept independently, and
expressed it in terms of the impossibility of continuously producing work by cooling
a system to a temperature below that of its surroundings. The two formulations have
been shown to be equivalent, however the discovery is usually credited to Clausius
who announced it earlier.
It is often convenient to write a balance equation [recall Eq. (4.10)] for entropy
dS
= Ṡ T + Ṡ G (5.28)
dt
in which the volume rate of entropy generation is
dS
Ṡ G = − Ṡ T (5.29)
dt
We can then cast the second law into a simple form in terms of this quantity that is
very useful.
Corollary 5.1 During an arbitrary processes undergone by a closed system the rate
of entropy generation is nonnegative.
Ṡ G ≥ 0 (5.30)
This can be interpreted in the following way. Although mass and energy can neither
be created nor destroyed, entropy can be created but not destroyed.
When a system undergoes an equilibrium process Ṡ G is zero. This can be deduced
from the fact that in this case the temperature T is uniform throughout the system.
Therefore, the entropy transfer, Eq. (5.26), is

1 Q̇
Ṡ =
T
−n · q̇ d A =
T A T
and is equal to d S/dt according to Axioms 3.2 and 5.1. Now an equilibrium process
is reversible, therefore, we are lead to the conception of Ṡ G as a measure of the process
irreversibility. A process with a larger value of Ṡ G is further from being reversible
than another with a smaller value of Ṡ G . As we have noted several times already,
an equilibrium process is not a real process. However, when a system undergoes a
quasi-equilibrium process, for which the system temperature is nearly uniform, and
Q̇ ∼ Q̇, then
d S/dt ∼ Q̇/T or Ṡ G ∼ 0
In other words, a quasi-equilibrium process is almost reversible.

The entropy generated in a process is the integral of the rate of entropy generation.
Therefore, from Eq. (5.30) we have
t2
S =
G
Ṡ G dt ≥ 0 (5.31)
t1
Equations (5.30) and (5.31) are useful when we have expressions for these
quantities.
5.4.2 Calculation of Entropy Generation
There are a few heat transfer situations for which we can easily evaluate Ṡ T and
thereby obtain [from Eq. (5.29)] a useable expression for Ṡ G
• when the heating is to or from a reservoir (see Sect. 3.3.3), R, at temperature TR ,
then even though the system, S, temperature varies, it is uniform and equal to TR
on the surface where the heating occurs. Thus we obtain
d SS Q̇ S/R
Ṡ SG = − (5.32)
dt TR
This is often integrated in time
Q S/R
S SG = SS2 − SS1 − (5.33)
TR
and applied to processes that take place between two equilibrium states.
• when there is heating to or from two reservoirs, a high temperature reservoir, H ,
whose temperature is TH , and a low-temperature reservoir whose temperature is
TL , the rate of entropy generation is
d SS Q̇ S/H Q̇ S/L
Ṡ SG = − − (5.34)
dt TH TL
and
Q S/H Q S/L
S SG = SS2 − SS1 − − (5.35)
TH TL
is the entropy generated in a process that takes place between two equilibrium
states.
• when the process is adiabatic there is no heating and consequently no entropy
transfer. Thus, Ṡ G is simply
d SS
Ṡ SG = (5.36)
dt
and
S SG = SS2 − SS1 (5.37)
for the entropy generated.

We can apply these considerations to specific problems.
Example 5.7 Calculate the entropy generated in the process of Example 5.2
considering the air to be a thermal reservoir at 70 ◦ F.
Solution.
In Example 5.2, we calculated the heat transfer to the coffee and the entropy
change of the coffee as
Q = −320.68 Btu S2 − S1 = −0.5149 Btu/R
Considering the air as a thermal reservoir allows use of Eq. (5.32) here
S G = S2 − S1 − Q/TR
Substituting numerical values gives
S G = −0.5149 Btu/R − (−320.68 Btu)/(529.67 R) = 0.0905 Btu/R > 0
This result indicates that the heat transfer process considered here is an irre-
versible process; it is possible and its reverse is impossible. Indeed Axiom 5.2
was established as a generalization of observations of this type.
The following example illustrates a mechanically irreversible process.
Example 5.8 Calculate the entropy generated per lbm when air considered
as a perfect gas is allowed to expand adiabatically in a free expansion process
from 100 psi and 500 ◦ F to twice its original volume.
Solution.
In this problem, only 7 equations have been written, therefore no extensive

information can be obtained. However all intensive information, for exam-
ple, the entropy generated per unit mass, can be determined. Now the spe-
cific volume ratio, v2 /v1 , is 2, and the temperature ratio, T2 /T1 is 1 (from
the energy equation since we are considering air as a perfect gas). There-
fore, the specific entropy difference as calculated using Eq. (5.21), s2 − s1 =
R ln[(T2 /T1 )1/(k−1) (v2 /v1 )], is
s2 − s1 = R ln 2
Since there is no heat transfer in this process we use Eq. (5.37) for the entropy
generated to obtain
S G /M = s2 − s1 = 36.97 ftlbf/lbmR > 0
In units more commonly used for entropy this is 0.0475 Btu/lbm R. Interest-
ingly we did not use the information given regarding the initial state, and we
could have done without it. This occurred because of the very simple expres-
sion for the entropic equation of state of a perfect gas. For a more complicated
class of substance, this would not happen. In any event, we find that this free
expansion process is irreversible.
The following example illustrates a process with both mechanical and thermal irre-
versibilities.
Example 5.9 One lbm of Neon is contained in a 4.5 ft3 cylinder at 90 psi
pressure by a piston that is pinned in place. After the piston is released, the
system comes to rest with the piston 1.5 times higher than it was initially with
a temperature, equal to the surrounding air, of 40 ◦ F. Assuming that the change
of state is polytropic, that Neon is a perfect gas, and that the air is a thermal
reservoir, how much entropy is generated in the process?
Solution.
State 1 State 2
M1 = p1 = Tf 2 = 40◦ F M2 = M1
1 lbm 90 psi (fixed mass)
V1 = V2 = 1.5V1
4.5 ft3 (geometry)
W = M R(T2 − T1 )/(1 − n) (polytropic-

ideal gas)
Q = U2 − U1 + W (energy
equation)
The initial temperature is obtained from the ideal gas equation of state

(90 psi)(4.5 ft 3 )/(1 lbm)(0.5316 psift 3 /lbmR)
T1 = p1 V1 /M1 R =
762 R (T f 1 = 302 ◦ F)
The polytropic exponent can be determined from
T1 v1n−1 = T2 v2n−1
by taking the natural logarithm
n − 1 = ln(762/500)/ ln 1.5 = 1.039
Then the work done by the Neon in the expansion is

(1 lbm)(76.55 ft lbf/lbm R)(762 − 500) R/1.039
W =
19300 ft lbf
The heat transfer during this process is
Q = (1 lbm)(0.15 Btu/lbm R)(40 − 302)R + (19300/778.2) Btu = −14.50 Btu
and the change in entropy is from Eq. (5.21)
s2 − s1 = R ln[(T2 /T1 )1/(k−1) (v2 /v1 )]
On substituting values into this, we get
s2 − s1 = (0.09837 Btu/lbm R) · ln[(500/762)1/0.67 1.5] = −0.02198 Btu/lbm R
The entropy generated is then from Eq. (5.33)
S G = M(s2 − s1 ) − Q/TR
and on using the appropriate values is
S G = (1 lbm)(−0.02198 Btu/lbm R) − (−14.5/500) Btu/R = 0.00702 Btu/R > 0
Therefore, this process is irreversible.
Thermally Isolated Systems

Problems involving heat transfer between systems, none of which are thermal reser-
voirs, but which together constitute a thermally isolated system also permit simple
evaluation of the rate of entropy generation. In a closed system, U , that is thermally
isolated from its surroundings, q̇ is zero over the entire surface. Consequently ṠUT is
zero, and Eq. (5.29) is
d SU
ṠUG = (5.38)
dt
Now if U is made up of a number, , of component subsystems, the additivity of S

allows us to write

SU = Sj
j=1
where S j is the entropy of the j th subsystem. The process is reversible if and only
if all the subsystems undergo equilibrium processes, namely, Ṡ Gj is zero for every
j. This is sometimes referred to as global reversibility. If at least one subsystem
undergoes a nonequilibrium process, the process is irreversible.
For processes that begin and end in equilibrium states, we get for the entropy
generated

SUG = (S j2 − S j1 ) (5.39)
j=1
The simplest application of this expression involves two systems, and therefore four
states.
Example 5.10 A 150 kg block of copper at 200 ◦ C is dropped into a bucket

of water whose volume is 0.17 m3 , and is at 25 ◦ C. The bucket is open to
the atmosphere, but no water vaporizes, and there is no heat transfer to the
surroundings. When the water–copper combination has reached a common
final temperature, how much entropy has been generated?
Solution.
The adiabatic condition is, on using Eq. (4.29), h 2 − h 1 = c p (T2 − T1 ) +

v O ( p2 − p1 ), for the enthalpy difference of both the copper and the liquid
water, and denoting the common final temperature by T2
MC c pC (T2 − TC1 ) + MW c pW (T2 − TW 1 ) = 0
With r = MW c pW /MC c pC the thermal mass ratio, the final Celsius tempera-
ture, Tc2 , is [this is Eq. (3.60) incorporating Eq. (1.49)]
Tc2 = TcW 1 + (TcC1 − TcW 1 )/(1 + r )
Using the given data, the thermal mass ratio is
(0.17 m3 )(1000 kg/m3 )(4.19 kJ/kg K)

r= = 12.50
(150 kg)(0.38 kJ/kg K)
so that the final temperature is
Tc2 = 25 ◦ C + [(200 − 25) ◦ C]/13.50 = 38.0 ◦ C
The entropy change of each component is calculated using Eq. (5.14)
s2 − s1 = c p ln(T2 /T1 )
so we get for the water and copper, respectively
sW 2 − sW 1 = 4.19 kJ/kg K ln(311.12/298.15) = 0.1784 kJ/kg K

sC2 − sC1 = 0.38 kJ/kg K ln(311.12/473.15) = −0.1593 kJ/kg K
Note that in this ratio you must use the absolute temperature, not the relative
one. The entropy generated by the thermally isolated system is therefore from
Eq. (5.39)
⎧
⎨ MW (sW 2 − sW 1 ) + MC (sC2 − sC1 )
SUG = (170 kg)(0.1784 kJ/kg K) + (150 kg)(−0.1593 kJ/kg K)
⎩
6.4 kJ/K > 0
This is an irreversible process.

5.4.3 Open Systems
In the previous chapter, we noted the importance of open systems for applications,
and developed a form of the first law for a fixed (control) volume. We wish to do
the same here for the second law, and we proceed in an entirely analogous fashion
(refer to Sect. 4.4). We start with a closed system consisting of a volume V , and N
feed pipes, denoted by S j , which exchange mass with the volume (see Fig. 5.2). The
closed system entropy is
N
S = SV + Sj
j=1
Differentiating and noting that the increase of entropy of the fluid in the j th feed
pipe, S j = s j M j , is [recall that d M j /dt = sign j Ṁ j , see Eq. (4.52)]
dSj dMj
= sj = sign j s j Ṁ j
dt dt
we find
dS d SV N
= + sign j s j Ṁ j
dt dt j=1
Applying the second law, Eq. (5.27), d S/dt ≥ Ṡ T , and noting, as we did in chapter 4
in connection with the energy equation, that there is no heat transfer through the feed
pipes (so that Ṡ Tj is zero), we obtain
d SV N
≥− sign j s j Ṁ j + ṠVT (5.40)
dt j=1
Fig. 5.2 Entropy increases in the volume V as a result of convection through the feed pipe j, and
transfer from the surroundings. The surroundings, S̃ consists of the air A, the piston P, and the
shaft S
We have already described the last term on the right of this inequality, the first term
is the entropy convected into the control volume. If we use the symbol ṠVC for this,
the entropy balance equation for the fixed volume has the descriptive form [recall
Eqs. (4.60)]
d SV
= ṠVC + ṠVT + ṠVG (5.41)
dt
The corresponding expression for ṠVG generalizes Eq. (5.29) to include convected
entropy
d SV
ṠVG = − ṠVC − ṠVT (5.42)
dt
However, the second law implication remains the same as for the closed system,
namely, that the rate of entropy generation is either zero for a reversible process, or
positive for an irreversible process. You can see this by substituting Eq. (5.40) into
Eq. (5.42).
Steady Flow Devices
Equation (5.42) can be applied to a steady flow device by setting SV equal to a
constant in similar fashion to Sect. 4.4.1. For a device with one inlet (denoted 1) and
one outlet (denoted 2), we obtain in this way

−nV · q̇
ṠVG = Ṁ(s2 − s1 ) − dA (5.43)
AV T
so that once we evaluate the integral (the entropy transfer to the volume), we have
the rate of entropy generation in the control volume. Just as with the closed system,
there are two special cases for which this is simple to do
• when the surroundings can be considered a thermal reservoir at temperature TR , the
surface temperature in the integral term of Eq. (5.43) is a constant TR everywhere
on the surface, so the integral is just the rate of heat transfer to the volume
ṠVG = Ṁ(s2 − s1 ) − Q̇ V /TR (5.44)
• in an adiabatic process there is no heat transfer to the control volume, hence there
is no entropy transfer and
ṠVG = Ṁ(s2 − s1 ) (5.45)
Example 5.11 The condenser of an air conditioning system inlets R-12 at

0.9 MPa, 50 ◦ C, and exhausts saturated liquid at 40 ◦ C. When the surrounding
air is at 30 ◦ C (consider this to be a reservoir), what is the rate of entropy
generation in the volume per lbm of flowing R-12?
Solution.
Substituting the steady flow energy equation into Eq. (5.44), we find for the
rate of entropy generation per unit mass rate of flow
ṠVG / Ṁ = (s2 − s1 ) − (h 2 − h 1 )/TR
Using the given state information, the values for the enthalpy and entropy in
state 1 are 211.765 kJ/kg and 0.7131 kJ/kg K, and in state 2 are 74.527 kJ/kg
and 0.2716 kJ/kg K respectively. As a result

ṠVG [(0.2716 − 0.7131) − (74.527 − 211.765)/(273.15 + 30)] kJ/kg K
=
Ṁ 0.01121 kJ/kg K > 0
for the entropy generated in the control volume.
5.5 Heating and Power Bounds
In the previous section, we described the second law of thermodynamics, and its
implication that the rate of entropy generation is not negative for all physical pro-
cesses that occur in closed and open thermodynamic systems. In the present section,
we describe some interesting deductions that can be made when the first and second
laws of thermodynamics are applied to simple generic systems composed of work
producing devices and thermal reservoirs.
5.5.1 Constraints on Heat Transfer
When a system undergoes a process while in contact with a thermal reservoir, the
second law imposes a constraint on the rate of heat transfer. This is an important
consideration because many practical devices are maintained in thermal contact with
either the atmosphere or another system that, due to its large mass, is effectively a
thermal reservoir (see Sect. 3.3.3).
Theorem 5.1 (Heating Bound) Let R be a thermal reservoir with temperature TR ,

and S a thermodynamic system. If R and S are each closed, and in thermal contact
only with each other, then the rate of heat transfer to S is bounded from above
d SS
Q̇ S ≤ TR ≡ Q̇ ∗S
dt
Proof The rate of entropy generation of the system S is given by Eq. (5.32)
d SS Q̇ S/R
Ṡ SG = −
dt TR
we rewrite this, noting that Q̇ S/R = Q̇ S because S is in thermal contact only with R
d SS
Q̇ S = TR − TR Ṡ SG
dt
The second law implies that Ṡ SG ≥ 0, and the definition of absolute temperature
requires TR ≥ 0. Therefore, the last term on the right in the equation above is not
negative, and since it is subtracted from the first, we find that
d SS
Q̇ S ≤ TR
dt
This is the result that was to be proved with Q̇ ∗S ≡ TR d SS /dt.
The quantity Q̇ ∗S , the heating bound, occurs when the process is reversible, and is
therefore called the reversible heating. Of course, it is the maximum possible value
of the rate of heat transfer to S. Several corollaries follow from this. The proofs of
these, which are short and simple, I leave to you.
Corollary 5.2 In an arbitrary process of a system S of the type described in Theo-

rem 5.1, the amount of heat transfer to the system is bounded from above.
Corollary 5.3 In an arbitrary process of a system S of the type described in The-

orem 5.1 that takes place between two equilibrium states, the heat transfer bound
is
Q ∗S = TR (SS2 − SS1 )
5.5 Heating and Power Bounds 205
Fig. 5.3 The system

considered in Theorem 5.2
where SS1 and SS2 are values of the system entropy in the initial and final states.
Corollary 5.4 In a steady-state process of a system S of the type described in The-
orem 5.1, the reversible heating is zero.
orem 5.1, heat is transferred from the system to the reservoir.
Corollary 5.5 indicates that a system that continuously transfers heat from a reservoir,
like the atmosphere, while maintaining a steady state is a contradiction of the second
law. This is true also for cyclic processes.
The results of this Theorem and its Corollaries can be applied to specific problems.
Example 5.12 Calculate the heating bound for the coffee cooling process
of Example 5.2 considering the air to be a thermal reservoir as described in
Example 5.7.
Solution.
The entropy change of the coffee was calculated in Example 5.2 as
S2 − S1 = −0.5149 Btu/R
and in Example 5.7 the air temperature was specified as 70 ◦ F. According to

Corollary 5.3, the heating bound is then
Q ∗S = (459.67 + 70) R(−0.5149) Btu/R = −272.7 Btu
This is indeed greater than the actual value, −320.7 Btu, calculated in
Example 5.2, as required by Corollary 5.2.
Another important process to which we can apply the second law, is the transfer of
heat between two reservoirs at different temperatures.
Theorem 5.2 (Clausius) Let H and L be thermal reservoirs with temperatures TH
and TL respectively, such that TH > TL . If the reservoirs are brought into thermal
contact by means of a closed, stationary system S, so that they exchange heat steadily,
then the heat transfer to S from H is bounded by
Q̇ S/H > 0
Proof The rate of entropy generation for this two reservoir process is given by
Eq. (5.34)
Ṡ SG = − −
dt TH TL
but this is a steady-state process for S; therefore, d SS /dt = 0 and
Q̇ S/H Q̇ S/L
Ṡ SG = − −
TH TL
Since S undergoes a steady-state process, the rate of increase of energy is also

zero, and there is no power developed because S is stationary. Therefore the energy
equation, Eq. (4.6)
dU S
= Q̇ S − Ẇ S
dt
applied to S, with dU S /dt = 0 and Ẇ S = 0, gives
Q̇ S = Q̇ S/H + Q̇ S/L = 0
Using this to eliminate Q̇ S/L in the expression for Ṡ SG produces

1 1
Ṡ SG = Q̇ S/H − (5.46)
TL TH
According to Eq. (5.30), Ṡ SG ≥ 0, and since the factor 1/TL − 1/TH is positive, the
factor Q̇ S/H must be either greater than or equal to zero. However Q̇ S/H = 0 by the
condition of steady heat transfer. Therefore Q̇ S/H > 0 as was to be shown.
Noting that in the steady heat transfer described in Theorem 5.2, Q̇ S/H > 0 based
on the existence of the hotness property (hot bodies heat cold bodies), we conclude
that in this situation the second law is consistent with the existing principles of heat
transfer and thermometry5 .
Let us now suppose that Q̇ S/H is given by Newton’s law of cooling, Eq. (3.48),
Q̇ S/H ≡ Q̇ L = h t A(TH − TL ). Substituting this into Eq. (5.46) above, we get
(TH − TL )2
Ṡ SG = h t A , (5.47)
T H TL
In this form the second law requires h t > 0, and is consistent with our previous
discussions of heating (see Sects. 1.3.9 and 3.3.1). According to the theorem, Q̇ S/H , is
positive (not zero) so that the heat transfer process is strictly irreversible. However let
5 Theorem 5.2 was adopted by Rudolf Clausius as an Axiom together with a variant of Axiom 5.1
and used to develop the second law and all of its consequences. It is known as the Clausius statement
of the second law of thermodynamics.
us consider what happens for small values of the temperature difference, TH − TL →

0. Generally, this implies that Q̇ S/H → 0, namely, that there is no heat transfer.
However, if simultaneously h t A → ∞, such that the heat transfer rate remains finite,
the entropy generation expression becomes, in this limit
2
1 Q̇ S/H
lim Ṡ SG = lim =0
h t A→∞ h t A→∞ h t A TH TL
This is heat transfer with zero entropy production (reversible heat transfer). It is
characterized by zero temperature difference and infinite surface area or heat transfer
coefficient.
Although reversible heat transfer is a fiction, as with other equilibrium processes,
we can conceive of heat transfer processes which are almost reversible. They are
facilitated by large area, A, and large heat transfer coefficient, h t . For them, as with
other quasi-equilibrium processes (see Sect. 5.4.1)
Ṡ SG ∼ 0
5.5.2 Constraints on Work
When we combine the results about heating bounds with the first law of thermody-
namics, we obtain corresponding results about power bounds. These are useful for
exactly the same reasons we stated before. Indeed they are the most useful results of
the second law for engineers.
Theorem 5.3 (Power Bound) Let R be a thermal reservoir with temperature TR ,
and S a thermodynamic system. If R and S are each closed, and in thermal contact
only with each other, then the power developed by S is bounded from above
Proof The first law, Eq. (4.6), for the system S has the forms
dU S
= Q̇ S − Ẇ S = Q̇ ∗S − Ẇ S∗
dt
Equating the second and the last of these
Q̇ S − Ẇ S = Q̇ ∗S − Ẇ S∗
and on rearranging
Ẇ S = Ẇ S∗ − ( Q̇ ∗S − Q̇ S )
From Theorem 5.1, the quantity in parentheses here is not negative, therefore we
have
Ẇ S ≤ Ẇ S∗
as was to be proved.
The quantity Ẇ S∗ , the power bound, occurs when the process is reversible, and is
therefore called the reversible power. Of course, it is the maximum possible value of
the rate at which S does work. As we did previously, we can obtain several corollaries
from this.
Corollary 5.6 In an arbitrary process of a system S of the type described in Theo-
rem 5.3, the amount of work done by the system is bounded from above.
Corollary 5.7 In an arbitrary process of a system S of the type described in The-
orem 5.3 that takes place between two equilibrium states, the bound on work done
is
W S∗ = U S1 − U S2 − TR (SS1 − SS2 )
where U S1 , SS1 and U S2 , SS2 are values of the system energy and entropy in the initial
and final states.
orem 5.3, the reversible power is zero.
Corollary 5.9 (Kelvin–Plank) In a steady-state process of a system S of the type
described in Theorem 5.3, work is done on the system.
Corollary 5.9 indicates that any process that continuously transfers heat from a reser-
voir, like the atmosphere, and simultaneously produces work while maintaining a
steady state, contradicts the second law6 . Similar results could be established for
cyclic processes. Ingenious inventors continually propose devices that are intended
to produce power by continuously extracting energy from the atmosphere or the
ocean. Such machines are called perpetual motion machines of the second kind
(refer to Sect. 4.2.1) because they supposedly never stop for fuel. No such machine
has ever worked as claimed, a fact consistent with the constraints discussed here.
Theorem 5.3 and its Corollaries can also be applied to specific problems.
Example 5.13 Calculate the reversible work for the released piston problem
described in Example 5.9
Solution.
In Example 5.9, the energy change and entropy change of the 1 lbm of Neon
contained in the cylinder were found to be
U2 − U1 = −39.3 Btu S2 − S1 = −0.02198 Btu/R
6 Corollary5.9 was adopted by both William Thomson (Lord Kelvin) and Max Plank as an Axiom
together with a variant of Axiom 5.1, and used to develop the second law and all of its consequences.
It is known as the Kelvin–Plank statement of the second law of thermodynamics.
Consequently, the reversible work associated with these changes while in con-
tact with a 40 ◦ F reservoir, is given by Corollary 5.7 as
W S∗ = U S1 − U S2 − TR (SS1 − SS2 )
or on substituting the appropriate values
W S∗ = 39.3 Btu − (500 R)(0.02198 Btu/R) = 28.31 Btu
This is 22000 ft lbf which is greater than the actual amount of work done, 19300
ft lbf, calculated in the example, as it should be according to Corollary 5.6.
The type of process described by Corollary 5.9 is one that is quite common. It
corresponds to frictional work being completely converted into heat such as occurs
when two blocks are continuously rubbed together. The dissipative nature of friction,
which has no counterpart in heat transfer, is the subject of the next theorem.
Theorem 5.4 (Dissipation) The irreversible part of the power developed by a closed
system of the type described in Theorem 5.3 is dissipative.
The power developed by a system, S, can be expressed in reversible and irreversible

components (see Sect. 3.2.2), and bounded as in Theorem 5.3
Ẇ S = Ẇ SR + Ẇ SI ≤ Ẇ S∗
where Ẇ SR (x, v) = −Ẇ SR (x, −v) and Ẇ SI (x, v) = Ẇ SI (x, −v). Then on substitut-
ing −v for v above (recall that this corresponds physically to carrying out the process
in reverse), the equation becomes
−Ẇ SR + Ẇ SI ≤ −Ẇ S∗
and adding gives

Ẇ SI ≤ 0
which means that Ẇ SI is dissipative.

Although we already knew this result from our experience (see Sect. 3.2.2), its
connection to the second law of thermodynamics is new. Indeed the power of the
second law is its ability to encompass such seemingly distinct concepts as heating
and friction.
Heat Engines
The analog of Clausius’ theorem for power in two reservoir systems has special
importance for engineering applications.
Theorem 5.5 (Carnot) Let H and L be thermal reservoirs with temperatures TH

and TL , respectively, such that TH > TL . If S is a closed system in thermal contact
Fig. 5.4 The system

considered in Theorem 5.5
with both H and L, which, on a continuous basis, does work at a rate Ẇ S while heat
is transferred to it from H at a rate of Q̇ S/H , then the power developed is bounded
by
TL
Ẇ S ≤ Q̇ S/H 1 −
TH
Proof For the system shown schematically in Fig. 5.4, the equations needed for the
analysis are, the rate of entropy generation, Eq. (5.34)
Ṡ SG = − −
dt TH TL
and conservation of energy for S,
dU S
= Q̇ S − Ẇ S ,
dt
where the rate of heat transfer to S is given by
Q̇ S = Q̇ S/H + Q̇ S/L
In this steady-state process, the energy and entropy of S remain constant so that the
energy equation and rate of entropy generation become, respectively,
Ẇ S = Q̇ S/H + Q̇ S/L (5.48)

Q̇ S/H Q̇ S/L
Ṡ SG = − − (5.49)
TH TL
Now on eliminating Q̇ S/L between these we obtain

TL
Ẇ S = Q̇ S/H 1− − TL Ṡ SG
TH
The second law implies that Ṡ SG ≥ 0, and the definition of absolute temperature
requires TL ≥ 0. Therefore, the last term on the right in the equation above is not
negative, and since it is subtracted from the first, we find that

TL
Ẇ S ≤ Q̇ S/H 1 −
TH
which was to be proved.
Clearly, a similar theorem could be proved for a system that operates on a cyclic
basis, and the result would be

TL
W S ≤ Q S/H 1 − (5.50)
TH
where the work done and heat transfer refer to one cycle. In either case, the maximum
continuous power output (work done per cycle), which occurs when the process is
reversible, is positive, not zero as it was for a system in contact with a single reservoir.
Such a system is a heat engine since it converts thermal power into mechanical power.
Indeed a major impetus for the development of thermodynamics was the invention
and improvement of steam engines at the beginning of the nineteenth century. The
key feature of these devices is that they convert heat into work on a cyclic basis.
Looking again at Fig. 5.1 reminds us how this is possible.
When we build an engine, we want to get the maximum output work for a unit
quantity of fuel burned, because the fuel expense increases with the amount we use.
Moreover, the amount of heat available for use in the engine also increases with the
amount of fuel consumed. All this means that a coefficient of performance, COP,
defined as the output work divided by the input heat transferred, is a useful measure
of comparison for heat engines; an engine with a larger COP will cost less to run at
the same power level as another with a smaller COP. From Eq. (5.50), this is simply
WS Ẇ S TL
COP ≡ ηT = = ≤1− ≤1 (5.51)
Q S/H Q̇ S/H TH
Since the COP cannot be greater than one, we are justified in calling it an efficiency,
the thermal efficiency of the engine. We see from the equation that the maximum
thermal efficiency of a heat engine that operates between two thermal reservoirs,
depends only on the absolute temperature of the reservoirs
TL
ηT∗ = 1 − (5.52)
TH
The corresponding maximum power is Ẇ S∗ = ηT∗ Q̇ S/H

A set of simple but useful problems can be solved using the definition of efficiency,
Eq. (5.51) along with the first law Eq. (5.48)
ηT = Ẇ S / Q̇ S/H Ẇ S = Q̇ S/H − Q̇ L/S
two of the four quantities ηT , Ẇ S , Q̇ S/H , Q̇ L/S are given and two can be determined
(recall a similar situation in connection with mechanical efficiency in chapter 3).
Other variations are also possible, for example, using the work and heat transfers per
cycle.
Example 5.14 Thermal pollution regulations at a certain site require that Q̇ L/S
be limited to 50 Btu/s. How much horsepower can be delivered at the site by
an engine whose thermal efficiency is 60%? If the site temperature is 30 ◦ F
and the engine source temperature is 900 ◦ F, what is the maximum horsepower
that could be delivered?
Solution.
This problem does not involve Q̇ S/H , therefore we eliminate it from the two
equations (note that Q̇ S/L = − Q̇ L/S )
Q̇ S/H = Ẇ S /ηT Q̇ S/H = Ẇ S + Q̇ L/S
and, in the present case, solve the remaining equation for Ẇ S
Ẇ S = Q̇ L/S /(1/ηT − 1)
Using the values provided in this equation produces
Ẇ S = [(50 Btu/s) · (1.415 hp s/Btu)]/(1/.6 − 1) = 106 hp
For the given source and site temperatures, the maximum efficiency is
ηT∗ = 1 − (459.67 + 30)/(459.67 + 900) = 0.64
so
Ẇ S∗ = [(50 Btu/s) · (1.415 hp s/Btu)]/(1/.64 − 1) = 126 hp
is the maximum power that could be delivered at the site.
5.5.3 The Carnot Cycle
The maximum efficiency is achieved by an engine–reservoir combination that is

reversible. It is called the Carnot efficiency in honor of Sadi Carnot who first studied
heat engines of this type, and described the processes that constitute the cycle. If
we consider the T, s plane and try to construct a cycle that encloses maximum area
(maximum work done) within the bounds of two reservoirs and a given Q S/H , it is
Fig. 5.5 A Carnot cycle as

viewed in a T, s plane. The
temperature of the working
fluid in each isothermal
process is infinitesimally
different from that of the
reservoir so that these are
reversible heat transfer
processes. The process as
shown proceeds clockwise
clear that the result will be the rectangular cycle shown in Fig. 5.5. This cycle, called a
Carnot cycle, consists of two isothermal heat transfer processes between the engine
and each reservoir, and two isentropic processes connecting them. The efficiency
of this cycle, which is the maximum possible for the two reservoirs, is given by
Eq. (5.52), and can be obtained geometrically by computing the area enclosed by
the cycle in Fig. 5.5 for W S and the area under the TH part of the cycle as Q S/H .
Moreover, this efficiency is the same for all reversible cycles that operate between
these two temperatures independent of the working fluid of the cycle or the substances
contained in the reservoirs.
Example 5.15 A methane–air mixture burns at a flame temperature of 2000 ◦ C.

What is the maximum efficiency of an engine operating between this temper-
ature and an ambient 25 ◦ C?
Solution.
Using Eq. (5.52) with the two given temperatures produces
ηT∗ = 1 − (25 + 273.15)/(2000 + 273.15) = 0.869
Therefore, a Carnot engine operating between these temperatures would con-

vert about 87% of the input thermal energy into mechanical (or electrical)
work.
Thermodynamic Temperature
A Carnot cycle is a reversible process so that there is zero entropy generated during
the course of a cycle. Accordingly, Eq. (5.35) applied over one cycle gives
Q S/H Q S/L
− − =0
TH TL
Accounting for the fact that Q S/L is negative, this can be written as
Fig. 5.6 A liquid–vapor Carnot cycle on the left, and a Clapeyron cycle on the right
Q S/H TH
= (5.53)
Q L/S TL
Comparing this with Eq. (5.4) shows that a Carnot cycle can be used to measure
thermodynamic temperature. The standard for this scale is the triple point of water
which has been defined as 273.16 K (459.69 R) (see footnote 14 in Sect. 1.3.9).
Therefore, temperatures above this value can be determined by measuring the values
of Q S/H and Q L/S in a Carnot cycle where TL = 273.16 K, TH is the temperature to
be determined, and
Q S/H
TH = · 273.16 K
Q L/S
For temperatures below the triple point, we would use a cycle that had the triple point
as the upper temperature.
Liquid–Vapor Carnot Cycles
Let us analyze the Carnot cycle shown in Fig. 5.6. The work done per cycle is
W S∗ = M(TH − TL )[sg (TH ) − s f (TH )]
and the heat transferred from the hot reservoir in each cycle is
Q S/H = M TH [sg (TH ) − s f (TH )]
The cycle that takes place between the same two temperatures, and consists of two
isothermal and two isochoric (constant volume) processes, the Clapeyron cycle, does
an amount of work per cycle
W S = M[ p H (TH ) − p L (TL )][vg (TH ) − v f (TH )]
which according to Theorem 5.5 is less than that of the Carnot cycle (because there
is irreversible heat transfer during the isochoric processes). Therefore
p H (TH ) − p L (TL ) sg (TH ) − s f (TH )

≤
T H − TL vg (TH ) − v f (TH )
In the limit TH − TL → 0, the difference ratio on the left-hand side becomes a

derivative, and the irreversibility vanishes because the isochoric processes vanish;
the inequality becomes an equality. As a result, we have
dps sg − s f
= (5.54)
dT vg − v f
where the derivative is the slope of ps (T ), the saturation curve for the liquid–vapor
transition. This equation, called Clapeyron’s equation after its discoverer, is important
because it allows us to obtain sg from s f without making a heat transfer measurement
[compare this with Eqs. (5.24) and (4.41) for the alternative].
Example 5.16 Use Clapeyron’s equation to calculate the value of sg at 100 ◦ F

in the steam tables.
Solution.
The tangent to the saturation curve is estimated most simply as the average of
the upper and lower chord slopes. These are from the tables at 100 ◦ F
(1.6927 − 0.94924)/(120 − 100) = 0.037173 psi/R
and
(0.94924 − 0.50683)/(100 − 80) = 0.0221205 psi/R,
respectively. The tangent is, therefore, 0.029647 psi/R, and the product of this
and the specific volume difference [refer to Eq. (5.54)] is
0.029647 lbf/in2 R · (350.4 − 0.016130) ft3 /lbm · 144 in2 /ft2 /(778.2 ft lbf/Btu)
which evaluates to 1.9222 Btu/lbm R. The saturated liquid entropy is obtained

from Eq. (5.14) using a constant c p value of 1, and s O = 0 at 32.018 ◦ F
s f (100 ◦ F) = ln[(100 + 459.67)/(32.018 + 459.67)] = 0.12950 Btu lbm R
This is exactly the value that appears in the tables, because the actual variation
with temperature of c p is so slight. The value of sg , on adding these results
as indicated in Eq. (5.54), is then 2.0517 Btu/lbm R. This is 3.5% larger than
the value that appears in the table, due to a less accurate evaluation here of the
tangent to the saturation curve than that in the table.
5.5.4 Refrigerators and Heat Pumps
A two reservoir heat engine operated in reverse transfers heat from the low-
temperature reservoir to the high temperature. We call this kind of device either
a refrigerator or a heat pump, depending on its use. If it is used to keep the low
temperature reservoir cool, it is a refrigerator or air conditioner, while if the purpose
is to keep the high-temperature reservoir warm, it is a heat pump.
Heat Pumps
For a heat pump, we are interested in Q̇ H/S > 0, therefore we write the result of
Carnot’s theorem as
TL
Ẇ S ≤ − Q̇ H/S 1 −
TH
According to conservation of power, Eq. (3.14), applied to this system, we have in

the absence of friction
Ẇ M + Ẇ S = 0
where Ẇ M is the power developed by the motor that is required to drive S. In a real
application, the motor must produce additional power to overcome dissipation in the
bearings (recall Sect. 3.2.4). Substituting this then gives

TL
Ẇ M ≥ Q̇ H/S 1− (5.55)
TH
A useful performance measure for a heat pump is the rate of heat transfer to the
warm space per unit power required for operation. From Eq. (5.54), this coefficient
of performance is
COP H = Q̇ H/S /Ẇ M ≤ 1/{1 − (TL /TH )} (5.56)
This is not an efficiency because it is greater than one. However, there is a largest
possible coefficient of performance, that of a Carnot heat pump, which is
COP∗H = 1/{1 − (TL /TH )} (5.57)
A typical heat pump arrangement is shown in Fig. 5.7. The heat pump is an environ-
mentally desirable way to heat a home because only Ẇ M units of energy are required
instead of Q̇ H/S which would be required if fuel were burned directly. Whether or
not it is profitable for a homeowner depends on the relative cost of electricity and of
fuel, however, in either case reducing the leakage, Q̇ L/H (improving the insulation),
is both environmentally and financially desirable.
Fig. 5.7 A model of a heat pump. The house can be considered as a reservoir because it is a steady-
state system, Q̇ H = 0. Note that the sum of Q̇ L/H and Q̇ L/S is not zero because there are other
(much larger) heating elements affecting the reservoir, L
Example 5.17 The electric bill for heating a house with a heat pump having
a COP of 3.75 was 90 dollars in November. If electricity cost 0.125 dollar/kW
hr, what was the average rate of heat loss from the house? If the house was
maintained at 22 ◦ C, and the average outdoor temperature was 1 ◦ C, how much
would it have cost to operate a Carnot heat pump?
Solution.
The average amount of electric power consumed was (there are 30 days in
November)
[(90 dollar)/(30 · 24 hr)]/(0.125 dollar/kW hr) = 1 kW
From Eq. (5.56), the rate of heat transfer to the house is then 3.75 kW, and
since this is equal to the rate of heat loss, this is the required value, in other
units more common to heat transfer it is 13500 kJ/hr. The COP of a Carnot
heat pump in the stated conditions is from Eq. (5.57)
COP∗H = 1/[1 − (273.15 + 1)/(273.15 + 22)] = 14.05
therefore, the average power required would be [from Eq. (5.56)] 3.75/14.05 kW.
On a monthly basis at the rate of 0.125 dollar/kW hr this is $24.02.
Fig. 5.8 A model of a refrigerator. The cold box can be considered as a reservoir because it is a
steady-state system, Q̇ L = 0. Note that the sum of Q̇ H/L and Q̇ H/S is not zero because there are
other (much larger) heating elements affecting the reservoir, H
Refrigerators
This case is sketched in Fig. 5.8. For a refrigerator, we are interested in Q̇ S/L , therefore
we eliminate Q̇ H/S between Eq. (5.55) for the heat pump and the energy equation
for this system
Ẇ M = Q̇ H/S − Q̇ S/L
This produces
TH
Ẇ M ≥ Q̇ S/L −1 (5.58)
TL
The minimum power required to drive a refrigerator operating between two reser-
voirs, is obtained when a Carnot cycle runs in reverse. For such a Carnot refrigerator,
the equality holds in Eq. (5.58).
A refrigerator coefficient of performance is defined similarly to that of a heat
pump
COP R = Q̇ S/L /Ẇ M ≤ 1/{(TH /TL ) − 1} = (TL /TH )/{1 − (TL /TH )} (5.59)
and again there is a largest possible coefficient of performance, that of a Carnot

refrigerator
COP∗R = 1/{(TH /TL ) − 1} = (TL /TH )/{1 − (TL /TH )} (5.60)
Problems involving refrigerators, air conditioners, and refrigeration systems in ther-

modynamics are similar to the engine problems we considered previously (for
instance Example 5.14).
Example 5.18 A Carnot refrigerator will be used to keep a cold box at −10 ◦ C,
and is to operate in an environment at 25 ◦ C. If the thermal leakage rate to the
cold box is 5 kW, what is the minimum power required to operate the cycle?
Solution.
The COP of the refrigerator is given by Eq. (5.60), and is
COP∗R = 1/{(273.16 + 25)/(273.16 − 10) − 1} = 7.519
The quantity Q̇ S/L is equal to the leakage rate for steady-state operation, there-
fore
∗
Ẇ M = Q̇ S/L /COP∗R = 5 kW/7.519 = 0.665 kW
An actual device would require more power to operate.
5.5.5 Thermodynamic Efficiency

We can use the restrictions placed on processes by the second law to define the
process efficiency. In this context, we are particularly interested in work producing
(or using) processes, and we make use of the Theorems and Corollaries of Sect. 5.5.2.
Two Reservoir Systems
We define the thermodynamic efficiency of a heat engine as
η = W S /W S∗
where W S∗ is the work produced by the Carnot engine working between the same two
reservoirs and transferring the same amount of heat

TL
W S∗ = Q S/H 1 −
TH
From Theorem 5.5, we find that η ≤ 1; it is truly an efficiency. Moreover, we can

express it in terms of the thermal and Carnot efficiencies defined earlier
η = ηT /ηT∗
Since this measure varies between zero and one, it gives an indication of the irre-
versibility present in the process; one indicates a reversible process. Moreover, it
is a better measure for irreversibility than the entropy generated because it is not
zero in a reversible process (with entropy generated we can compare two irreversible
processes and tell which is more irreversible, but we cannot compare an irreversible
process with a reversible one as we can with thermodynamic efficiency).
Fig. 5.9 When a system is

in thermal contact with a
single reservoir, the process
that connects states 1 and 2,
and produces the maximum
amount of work, consists of
reversible isothermal and
isentropic parts
Example 5.19 Calculate the thermodynamic efficiency of the engine reservoir

system described in Example 5.14.
Solution.
In Example 5.14, the thermal efficiency of the site engine is 60%, and the max-
imum thermal efficiency at the site is 63.99%. Therefore, the thermodynamic
efficiency is
η = 0.60/0.6399 = 0.9376
or approximately 94% efficient.
In a similar vein, we define the thermodynamic efficiency of a refrigerator or heat

pump as
∗
η = WM /W M
∗
where W M is the work required to drive a Carnot refrigerator or heat pump, respec-
tively. Consequently η ≤ 1, and can be expressed in terms of the COP as
η = COP R / COP∗R η = COP H / COP∗H
Single Reservoir Systems

For processes involving systems with a single reservoir, Theorem 5.3 with Corol-
lary 5.4 provides the appropriate definition of thermodynamic efficiency
η = W S /W S∗ WS > 0 or η = W S∗ /W S WS < 0
The reversible work, W S∗ , is given in terms of the change of state by Corollary 5.7;
W S∗ = (U S1 − U S2 ) − TR (SS1 − SS2 ). The reversible process that produces the max-
imum amount of work, W S∗ , in the given change of state, is a piece of the Carnot
cycle that passes through the two state points and exchanges heat with the reservoir
at TR , as shown in Fig. 5.9.
Example 5.20 Calculate the thermodynamic efficiency of the released piston

problem described in Examples 5.9 and 5.13.
Solution.
The actual work done in the process as calculated in Example 5.9 is 19300
ft lbf, while the reversible work calculated in Example 5.13 is 21000 ft lbf.
Consequently the thermodynamic efficiency of the process is
η = 19300/21000 = 0.919
or approximately 92% efficient.
For processes which are intended to transfer heat, we need another measure. For
these heat exchangers we use the results of Sect. 5.5.1 and define the thermodynamic
effectiveness as
e = Q S /Q ∗S QS > 0 or e = Q ∗S /Q S QS < 0
This is less than or equal to one according to Corollary 5.2, and Q ∗S is given by
Corollary 5.3, Q ∗S = TR (SS2 − SS1 ).
Example 5.21 Calculate the thermodynamic effectiveness of the heat transfer

problem described in Examples 5.2 and 5.12.
Solution.
In Example 5.2, the actual heat transfer to the air was calculated as −320.7
Btu, and in Example 5.12 the reversible heat transfer was calculated as −272.7
Btu. Therefore, the thermodynamic effectiveness for this process is
e = (−272.7)/(−320.7) = 0.8503
or approximately 85% effective.
Open Systems
In the case of an open system, the results of Sect. 5.5.2 cannot be used. However,
analogous results can be obtained. These are most interesting for the case of one
inlet, one outlet, steady flow devices that are in thermal contact with the atmosphere.
Turbines and compressors are practical examples to which such analysis applies.
Here we use the steady flow energy equation, Eq. (4.62), with the approximations
applicable to a gas flow
Ṁ(h 2 − h 1 ) = Q̇ − Ẇ
and the entropy generation expression, Eq. (5.44), with the second law incorporated
Q̇ ≤ Ṁ TR (s2 − s1 )
Solving the first of these for Ẇ and eliminating Q̇ by means of the inequality produces
the desired result
Ẇ / Ṁ = w ≤ h 1 − h 2 − TR (s1 − s2 )
This means that in a steady flow through a two feed pipe device the work done (power
developed) is bounded by the reversible work done
Ẇ ∗ / Ṁ = w ∗ = h 1 − h 2 − TR (s1 − s2 )
Therefore, we define a thermodynamic efficiency as before
η = w/w ∗ w>0 or η = w ∗ /w w<0
As you can see, the sole difference between the open and closed system cases is that
either enthalpy or internal energy appears in the maximum work expression.
Example 5.22 Steam expands in a turbine from an initial pressure of 100 psi
and temperature of 700 ◦ F, to the saturated vapor state at 5 psi. If the heat
transfer to the atmosphere is measured to be 18 Btu/lbm, and the atmosphere
can be considered a reservoir at 60 ◦ F, calculate the work done per pound of
steam, and the thermodynamic efficiency of the turbine.
Solution.
From the tabulated values, we find in state 1
h 1 = 1379.2 Btu/lbm s1 = 1.8033 Btu/lbmR
and in state 2
h 2 = 1131 Btu/lbm s2 = 1.8441 Btu/lbmR
The work done per lbm is then

Fig. 5.10 A geometrical interpretation of the turbine and compressor efficiencies on the left and
right, respectively

(1379.2 − 1131 − 18) Btu/lbm
w=
230.2 Btu/lbm
while the reversible work done is

[(1379.2 − 1131) + 519.67(1.8441 − 1.8033)] Btu/lbm
w∗ =
269.4 Btu/lbm
The thermodynamic efficiency is, therefore, 85.4%.
An alternate definition of efficiency is often used for compressors and turbines,

primarily because it is easy to implement. The definition has the form η = w/w∗ as
before, however the assumption is made that the device is adiabatic (this is consistent
with the second law provided that s2 ≥ s1 ). This means that w = h 1 − h 2 and w ∗ =
h 1 − h 2s , where state 2s is a fictitious isentropic final state. This fictitious state
is uniquely specified by making its pressure the same as the actual final pressure.
Although this procedure is somewhat ad hoc, it has the advantage that it does not
require knowledge of the reservoir temperature or the amount of heat transfer. Indeed
the calculated amount of work is in error by the amount of heat transfer. The turbine
efficiency is
η = (h 1 − h 2 )/(h 1 − h 2s ) (5.61)
while for a compressor

η = (h 1 − h 2s )/(h 1 − h 2 ) (5.62)
since the actual work required to drive the compressor is greater than the reversible
work. These relations are illustrated on the h, s diagram curves of Fig. 5.10.
Example 5.23 Steam expands in a turbine from an initial pressure of 100 psi
and temperature of 700 ◦ F, to a saturated vapor state at 5 psi. Calculate the
turbine efficiency.
Solution.
The state 1 and 2 conditions are the same as in the previous problem
h 1 = 1379.2 Btu/lbm s1 = 1.8033 Btu/lbmR
and
h 2 = 1131 Btu/lbm s2 = 1.8441 Btu/lbmR
Therefore, the work done per lbm is with the adiabatic assumption, 248.2
Btu/lbm (this is about 8% greater than calculated in the previous example.
State 2s is a liquid–vapor equilibrium because the entropy is less than the
saturated vapor entropy at 5 psi. Consequently, the quality can be calculated
from Eq. (5.25) using data from the saturation curve table
x2s = (1.8033 − 0.23486)/1.60924 = 0.9746
the fictitious final enthalpy then follows from this
h 2s = 130.17 + 0.9746 · 1000.83 = 1105.6 Btu/lbm
and the fictitious work done per lbm is
w ∗ = (1379.2 − 1105.6) Btu/lbm = 273.6 Btu/lbm
The turbine efficiency is then 84.1%.
As we have seen, the method of applying the first and second laws to the interac-
tions of finite size systems yield easy to obtain gross estimates of performance. That
is what makes thermodynamic analysis important in engineering practice.
5.6 Exercises 225
5.6 Exercises
Section 5.2
5.1 In a triangular cycle on a T, s plane, consisting of an isentropic expansion from
TH to TL < TH at s R , followed by an isothermal heat rejection at TL from s R to
s L < s R , followed by a straight line process in this plane back to the starting point,
derive expressions for the heat transfer on the diagonal line, and the net work done
in this cycle.
5.2 Water is heated in a quasi-equilibrium process T = T1 (s/s1 )n (in this expression

the T is the absolute temperature) from the saturated liquid state at 14.7 psi to the
saturated vapor state at 360 ◦ F. What is the value of n and how much heat is absorbed
per lbm of water in this process? (0.123, 957 Btu/lbm).
Section 5.3
5.3 Use Eq. (5.6) and Eq. (2.27) to eliminate Dt T p and thereby deduce the third form
of the entropic equation of state in terms of p and v. Use this to obtain the partial
derivatives ∂v s| p = c p /(T vα) and ∂ p s|v = c p vβ/(T α). These are the remaining
two Maxwell relations.
5.4 Use the result of the previous problem to derive the result ∂ p v|s = −cv vβ/c p .
Hint: Use one of Eq. (2.26).
5.5 Use the result of the previous problem to derive the result ∂ρ p|s = c p v/(cv β),
where ρ = 1/v is density. Hint: Use the other one of Eq. (2.26).
5.6 The quantity ∂ρ p|s has dimensions velocity squared and is the square of the
speed of√
sound, c. Use the result of the previous problem to show that for an ideal
gas c = k RT , where k is the ratio of specific heats.
Section 5.3.1
5.7 Glycerin exists at atmospheric pressure (101 kPa) and a temperature of 18 ◦ C.
If it is stirred, in a vat open to the atmosphere, until its temperature is 46 ◦ C, what is
the change in its specific entropy?
5.8 A 1.5 hp motor which has an efficiency of 92% is used to stir 5 lbm of machine oil
initially at 70 ◦ F. If after 5 minutes the motor is switched off, the process undergone
by the oil was adiabatic, and assuming that no oil is lost in the process, what is the
final temperature, and what is the change in entropy? (216 ◦ F, 0.488 Btu/R).
5.9 A stream of cold water at 40 ◦ F, 20 psi is mixed with a hot stream at 160 ◦ F, 40
psi so that the outlet stream is at 100 ◦ F, 14.7 psi. What is the change in entropy in
the mixing chamber per lbm of cold water flowing in this steady flow? What is the
increase in entropy in the surroundings per lbm of cold water flowing?
5.10 A 1.5 hp motor which has an efficiency of 92% is used to stir 5 lbm of water
contained in an open vessel, initially at 70 ◦ F. When the motor is switched off and the
water has come to rest its temperature has just reached the boiling point. Assuming
that no water has boiled away, that the process is adiabatic, and that the atmospheric
pressure is 14.7 psi, what is the change in entropy? How long did the motor run?
(1.188 Btu/R, 12.1 min).
Section 5.3.2
5.11 Show, by integrating Eq. (5.1) using Eq. (5.21), that the heat transfer in an
equilibrium isochoric process of a perfect gas is given by
Q = Mcv (T2 − T1 )
5.12 In a duct flow of a perfect gas, at the stagnation location (where the area is
infinitely large and the velocity is zero), the pressure is p0 , the temperature is T0 , and
the mass rate of flow is Ṁ. At another section downstream, the temperature is T , and
the flow between these locations is isentropic and steady. Based on this information,
derive the expression for the downstream area, A2 = A
1/(k−1) −1
A T T 1/2
= 1−
AC T0 T0
where AC = Ṁ(k − 1)1/2 (RT0 )1/2 /[(2k)1/2 p0 ] is a characteristic area, and the
expression for the downstream velocity, v2 = v, is
1/2 1/2
2 T
v = c0 1−
k−1 T0
where c0 = (k RT0 )1/2 is the speed of sound at the stagnation location.
5.13 Plot the results of the previous problem A/AC and v/c0 as well as the ratio of
the local speed of sound c = (k RT )1/2 to c0 as functions of 0 ≤ T /T0 ≤ 1 for air
(k = 1.4).
5.14 Show by differentiation that the area is minimum when T /T0 is equal to 2/(k +
1). Also show by substitution that at this section the velocity is equal to the local
speed of sound. At this point the flow is sonic; when v < c the flow is subsonic, and
when v > c the flow is supersonic.
5.15 Air flows steadily through an isentropic turbine, entering at 150 psi, 900 ◦ F,
and 350 ft/s, and leaving at 20 psi, and 700 ft/s. If the inlet area of the turbine is 0.1
ft2 , what is the outlet area, the mass rate of flow, and the power output? Treat the air
as an ideal gas.
5.6 Exercises 227
5.16 R-12 enters an isentropic compressor at 14.7 psi, 20 ◦ F and a volume flow
rate of 10 ft3 /s. If the final pressure is 100 psi, what is the mass flow rate, the final
temperature, and the power input? (3.55 lbm/s, 147 ◦ F, 81.0 hp)
at 50 kPa. If the flow is isentropic, the inlet area is 150 cm2 , and the exit area is 820
cm2 , what is the mass flow rate, the exit velocity, and the power output?
5.18 Nitrogen is compressed from 14.7 psi, 60 ◦ F to a pressure of 150 psi, in an

isentropic process. If the volume rate of flow at the inlet is 500 ft3 /min, what is the
mass rate of flow, the final temperature, and the power input to the compressor?
(0.615 lbm/s, 550 ◦ F, 106 hp)
5.19 R-12 enters a diffuser as saturated vapor at 30 ◦ C with a velocity of 120 m/s,
and leaves at 40 ◦ C. If the flow is isentropic, and the exit area is 0.001 m2 , what is
the mass rate of flow, and the exit velocity?
Section 5.3.3
5.20 Steam enters an isentropic turbine at 1250 psi, 800 ◦ F and leaves at 5 psi. What
is the final quality, and what mass flow rate is required for a power output of 1 MW?
(80.9%, 2.17 lbm/s).
5.21 One and one half lbm of water contained in a rigid vessel at 80 ◦ F is heated
until it exists as saturated vapor at 3000 psi. What is the change in entropy of the
water?
Section 5.4.2
5.22 Steam is contained in a 1.7 ft3 tank at 600 psi and 500 ◦ F. A valve is opened
and the steam expands into a larger volume. If the final pressure is 120 psi, the final
temperature is 360 ◦ F, and the heat transfer is to a reservoir at 250 ◦ F, how much
entropy is generated by the steam? (0.340 Btu/R)
5.23 Steam is contained in a 1.7 ft3 tank at 600 psi and 500 ◦ F. A valve is opened
and the steam expands into a larger volume. If the final pressure is 120 psi and the
process is adiabatic, how much entropy is generated by the steam?
5.24 Helium is compressed in a polytropic process, with n = 1.42, from 14.7 psi and
70 ◦ F to half its original volume. What is the work done by the helium, and the heat
transfer to the helium. If the heat transfer is to a reservoir at the initial temperature,
calculate the entropy generated. (−211 Btu/lbm, −78.9 Btu/lbm, 0.0189 Btu/lbm R)
5.25 Show that in an isothermal process of a perfect gas in contact only with a
thermal reservoir at temperature, TR , if v2 > v1 then TR > T and if v2 < v1 then
TR < T .
Section 5.4.3
5.26 Air enters a nozzle at 50 psi, 150 ◦ F, 150 ft/s, and leaves at 14.7 psi, 900 ft/s.
The heat loss from the nozzle is 6.5 Btu/lbm of air flowing, to a 100 ◦ F reservoir,
and the inlet area is 0.1 ft2 . What is the rate of entropy generation?
5.27 Cold water at 40 ◦ F and 60 psi is heated in a chamber by mixing it with saturated
vapor at 60 psi. If both streams enter the chamber with the same mass flow rate, and
the pressure of the exiting stream is also 60 psi, what is the rate of entropy generation
per lbm cold water flowing? (0.0735 Btu/lbm R)
5.28 A stream of hot air is cooled from 100 ◦ F, 20psi to 70 ◦ F, 15 psi by passing
it around a tube bank filled with flowing cold water. If the water enters at 40 ◦ F,
20 psi and exits at 60 ◦ F, 14.7 psi, if the required air flow rate is 3 lbm/min, and if
the overall system is adiabatic, what is the rate of entropy generation? At these low
temperatures, you may consider air to be a perfect gas.
5.29 R-12 at 120 psi, 100 ◦ F is throttled to 80 ◦ F. How much entropy is generated
per lbm of R-12 flowing? (0.01376 Btu/lbm R)
at 50 kPa, 150 ◦ C. If the flow is adiabatic, the inlet area is 150 cm2 , and the exit area
is 820 cm2 , what is the mass flow rate, the exit velocity, the power output, and the
rate of entropy generation?
5.31 Air is compressed from 14.7 psi and 70 ◦ F to a pressure of 100 psi and 600 ◦
F. If the work required to compress the air is 130 Btu/lbm and the heat transfer is to
a reservoir at the initial temperature, how much heat is transferred to the air and how
much entropy is generated per lbm? (-37.2 Btu/lbm, 0.3703 Btu/lbm R)
Section 5.5.2
5.32 Construct proofs for Corollaries 5.2 through 5.9.
5.33 A heat engine receives energy from a solar collector at a temperature of 175 ◦ F,
and rejects energy to its surroundings at 80 ◦ F. The solar collector converts 50% of the
incident energy to useable thermal energy. If 300 Btu/hr of solar energy illuminates
each ft2 of collectors, what is the minimum area required to provide 7 hp of power?
(794 ft2 )
5.34 The prime movers used in a nuclear power plant are heated by 5 · 109 Btu/hr
from the core at 620 ◦ F, and reject thermal waste to a nearby river at 80 ◦ F. What is
the maximum possible plant efficiency, and corresponding maximum power output?
5.35 A steam power plant is constructed which produces 75 kW of power while

rejecting 190 kJ/s to the condenser. How much heat is transferred from the boiler,
and what is the thermal efficiency? (265 kW, 28.3%)
5.6 Exercises 229
Section 5.5.4
5.36 A refrigerator operates in a room which has a temperature of 21 ◦ C. It must
reject 75 kW to the room in order to maintain the cold space at −10 ◦ C. What is the
minimum size (hp) motor required to satisfy these specifications? (10.6 hp)
5.37 A room air conditioner extracts 3000 Btu/hr while dumping 4500 Btu/hr out-
side. What is its COP, and how many amperes will it draw from a 110 v circuit? If the
room is at 75 ◦ F, and the outside temperature is 95 ◦ F what is the COP of a Carnot
air conditioner?
5.38 A dorm is to be maintained at 22 ◦ C by means of a heat pump pumping from

the atmosphere. If the value of h t A [see Eq. (3.48)] is 5 kJ/s K in the expression for
the rate of heat loss from the dorm, and the outside temperature is −5 ◦ C, what is the
minimum power required to drive the pump? If the device is used in the summer to
cool the dorm with the same interior temperature, the same power, but 80% of the
value of h t A, what is the maximum outside temperature? (12.3 kW, 52.1 ◦ C)
Section 5.5.5
5.39 One and a half lbm of argon is heated slowly in a piston-cylinder arrangement
from 30 psi and 300 ◦ F until the height increases by 30% and the pressure increases
by 50%. If the process is linear and the source is a reservoir at 1100 ◦ F, what are the
heat transfer and work bounds, and the thermodynamic efficiency and effectiveness?
(150 Btu, 67.7 Btu, 31.4%, 69.1%)
5.40 One and a half lbm of steam is heated slowly in a piston-cylinder arrangement
from 40 psi, 320 ◦ F until the temperature increases to 500 ◦ F. If this is a constant
pressure process and the source is a reservoir at 1100 ◦ F, what are the heat transfer
and work bounds, and the thermodynamic efficiency and effectiveness?
5.41 Saturated water vapor at 200 ◦ C is contained in a vessel whose volume is 25 l.

If the steam undergoes a free expansion to a pressure of 200kPa at 200 ◦ C, what is
the work bound, how much heat is transferred to the water from a 200 ◦ C reservoir,
what is the heat transfer bound, and the thermodynamic effectiveness? (88.1 kJ, 11.6
kJ, 99.7 kJ, 11.6%)
5.42 One kg of R-12 is cooled in a cylinder from 1 MPa, 110 ◦ C to a final temperature
of 60 ◦ C. Assuming that the process is isobaric and that the surroundings are a
reservoir at 50 ◦ C, how much work is done by the R-12, how much heat is transferred
to it, what are the work and heat transfer bounds, and the thermodynamic efficiency
and effectiveness?
at 50 kPa, 150 ◦ C. If the flow is adiabatic, the inlet area is 150 cm2 , and the exit area
is 820 cm2 , what is the turbine efficiency? (91.1%)
Chapter 6
Power and Refrigeration
6.1 Introduction
It is impossible to overestimate the impact that the development of steam- and

gasoline-powered engines have had on the history of western civilization. Indeed, the
availability of large sources of power fueled the industrial revolution, while sources
of increasingly higher power density extended the range of travel to the global scale.
Today we are so dependent on easy access to power that it is difficult to imagine life
without it. In this part of the current chapter we will study the technical aspects of
this development.
6.2 Vapor Power Cycles
The earliest development was the use of water vapor, steam, to create a source of
power that was unmatched by anything that was available until then, although at the
outset its utility was quite limited.
6.2.1 The Newcomen Engine
Thomas Newcomen put the first commercially successful heat engine into operation
in 1712 (some 300 years ago); it was used to drive a pump that drained water from
a coal mine. A schematic diagram of it is shown in Fig. 6.1. The Newcomen engine
operated on the principle that when steam at one atmosphere pressure is condensed
by reducing its temperature, the condensation pressure, which depends on the tem-
perature, is less than one atmosphere. The resulting pressure differential (the other
side of the piston has atmospheric air pushing down on it generates a force large
enough to drive the piston down.

232 6 Power and Refrigeration
Fig. 6.1 Drawing of a Newcomen engine connected to a water pump (From Wikipedia Commons)
Thus, spraying cold water into a cylinder filled with steam causes the steam in
it to condense and thereby creates a pressure differential that begins to drive the
piston down when it is equal to the external load divided by the piston area (this
occurs at point 2, pressure pb , in Fig. 6.2, note that the outer surface of the piston
has atmospheric air, pressure pa , acting on it). In actuality, the pressure drops below
the point 2 pressure as the piston moves down and condensation continues, but for
the sake of simplicity we assume that the process between points 2 and 3 is isobaric
with pressure pb . This is a crude approximation but it allows a simple analysis, and
the object is not an accurate analytic description of the Newcomen engine, but rather
a result to which we can compare James Watt’s subsequent improvements. When
the piston reaches the bottom of the cylinder (point 3 in Fig. 6.2) valves are opened
that let the condensed water drain out and saturated steam from a boiler enter the
cylinder at atmospheric pressure, pa ; then as the pressure in the cylinder returns to
atmospheric (point 4 of the figure) the weight of the external load begins to pull the
piston up. The drain valve is closed and as the piston rises steam fills the cylinder.
Here we have made another, similar, simplification in which we approximate the
process by an isochoric increase in pressure followed by an isobaric increase in
volume. The reason for this is the same as before. We accept less accuracy of result
in favor of simplicity in calculation (engineers find that making an idealized analysis
6.2 Vapor Power Cycles 233
Fig. 6.2 The idealized

cylinder pressure in a
Newcomen engine as a
function of the cylinder
volume
of an engine and its cycle is a fruitful way to understand it, and you will see this
process repeated later with internal combustion engines). When the piston reaches
the top of the cylinder (point 1 of the figure), a valve that allows cold water to spray
into the cylinder is opened and the cycle repeats. The valves are opened and closed
automatically by means of control rods connected to the walking beam.1
Because the atmospheric air pressure is the source of the force that drives the
piston down, the Newcomen engine was also called an atmospheric engine.
Engine Power
The engine has two elements, each with a distinct function: a boiler that converts
energy from chemical to thermal form by burning fuel, and transfers the energy
to water by boiling it, and a prime mover, the piston in cylinder, that converts the
thermal energy into mechanical energy of motion. The work done by the steam in a
cycle of the prime mover can be calculated from Fig. 6.2 as
W = ( pa − pb )(Vb − Vc ) = pmep VD (6.1)
where pmep is the mean effective pressure, and VD = A p L S is the stroke volume or
displacement of the piston. The average power developed is then
{Ẇ } = W/tc = W N (6.2)
where N is the number of cycles per unit time. These quantities relate to the engine
output. Some of this goes toward doing useful work while some of it is dissipated by
friction as we discussed in Chap. 3. As we learned there, the mechanical efficiency
measures the fraction of useful work.
The following example illustrates some of these quantitative aspects of prime
mover operation.
1 Thename of this part of the engine derived from an earlier version of the pump in which a man
walked back and forth along the beam which correspondingly rose and fell, thereby producing the
pumping action.
Example 6.1 The original atmospheric engine had a 21 inch diameter cylinder
and a 7 ft stroke. With an external load, Fe , of 2700 lbf and engine speed of 12
cyc/min, how much power did the engine produce? The engine was connected
to a pump that raised 10 gal of water 150 ft on each power stroke, what was
its mechanical efficiency?
Solution.
A pressure stroke diagram for the engine is shown at the right. At point 1, the
steam inlet valve is closed and the water spray valve is opened. As the water
spray lowers the temperature, the corresponding saturation pressure decreases
and when the pressure reaches
p2 = pa − Fe /A p
the pressure differential across the piston causes it to move down to the
bottom of the cylinder. Consequently, with the piston area A p = π(10.5 in)2 ,
the mean effective pressure is
pmep = Fe /A p = 2700 lbf/(346.36 in2 )
or 7.79 psi (see the accompanying figure). The work done by the engine is
given by Eq. (6.1)
⎧
⎨ ( pa − pb )VD = ( pa − p2 )L S A p = Fe L S
W = (2700 lbf)(7 ft)
⎩
18900 ft lbf
the cycle time, tc = cycle/N , is 1/12 minute, or 5 s, so the power developed

is, from Eq. (6.2)
Ẇ = (18900 ft lbf)/[(5 s)(550ft lbf/hp s)]

or 6.87 hp. Since the pump raised 10 gal of water 150 ft on each stroke the
work required to operate it is
⎧
⎪
⎪ V gh/v O
⎨
(10 gal)(0.1337 ft3 / gal)(32.174 ft/s2 )(150 ft)/
WP =
⎪
⎪ [(.016 ft3 / lbm)(32.174 lbm ft/lbfs 2 )]
⎩
12530 ft lbf
This is the negative of the reversible work done by the surroundings so that
the mechanical efficiency of the engine is simply
η M = 12530/18900 = 0.66
which is 66% efficient.
The low value of the mechanical efficiency calculated here for the Newcomen
engine was due to the primitive state of mechanical craftsmanship at that time. For
example, boring an 8 ft long hole in an iron casting, straight enough not to bind
the piston in its stroke and accurate enough not to allow an excessive amount of
steam to leak around the piston, was a task that represented the then existing limit of
manufacturing technology.
Engine Characteristics
The amount of power that could be produced by a Newcomen engine far exceeded
that of other contemporary sources: domestic animals, men, water, and wind. This
fact lead to their increasing utilization by industrialists in the early eighteenth century.
However, two features of these devices prevented their complete commercial success:
their reciprocating motion output, and their low efficiency of energy conversion. The
reciprocating output motion was not suitable in circumstances that required rotary
motion, such as milling grain and driving machinery. Moreover, the cost of the coal
that had to be burned as fuel in order to convert the energy made the use of these
engines prohibitively expensive for many, otherwise power needy, applications.
The reciprocating nature of the motion affects the relationship between the time
to complete a cycle, tc (recall that the engine running speed is N / cycle = 1/tc ), and
the applied load. For example, if the applied load is larger than pa A p + m p g (where
m p g is the piston weight), the piston does not move at all; and as the load decreases
from this limiting value, the speed increases until it reaches an ideal maximum value
when there is no applied load. A graph of the net load (load minus piston weight)
is shown in Fig. 6.3 along with one of the work done by the net load as a function
of speed and the average power developed as a function of speed. Interestingly, the
last curve has a maximum value, and consequently, the engine should be operated
near the corresponding speed. For the Newcomen engine, this operating speed can
be obtained by adjusting the net external applied load, Fe , as shown in Fig. 6.3.
Fig. 6.3 The force–speed, work–speed, and power–speed curves for a reciprocating engine
Engine Economy
Of two engines that use fuel at the same rate, the economically more desirable one
produces more power, since that one makes more power available for the same cost.
Thus, if c is the price of a unit mass of fuel, the ratio c{ Ṁ f }/{Ẇ }, the cost to produce
a unit amount of power, is an important quantity for making economic decisions
regarding the use of engines. Consequently, the specific fuel consumption
sfc = { Ṁ f }/{Ẇ } (6.3)
is a useful measure of the performance of engines; the smaller the sfc of an engine,
the more economical it is to operate (because less fuel is required to produce a unit
amount of energy). For the Newcomen engine as well as others that burn fuel to
create steam and then use the steam to produce mechanical work, the sfc is written
as
{ Ṁ f } 1
sfc = (6.4)
{ Ṁ} w
where { Ṁ} is the average mass flow rate of steam. The first factor of this expression
depends on the boiler and the second factor on the prime mover. This discussion
shows that in an economic sense, the specific work done by the steam, w = W/M1 ,
is a more important parameter than either W or {Ẇ }. Clearly, two prime movers that
are provided with steam under the same conditions can be compared on the basis of
their values of w, and the one with a larger w is more economical to operate. The
following example illustrates this calculation for the original Newcomen engine.
Example 6.2 The engine of Example 6.1 has a clearance, L c of 0.5 ft. Steam
enters the cylinder from the boiler as saturated vapor at 14.696 psi as the piston
moves up. Assuming that the cylinder is empty in state 3, calculate the work
done per unit mass of steam.
Solution.
The mass of steam that enters the cylinder in each cycle is M = V1 /v1 . With
a cylinder volume of
V1 = A p (L c + L S ) = [(346.36/144) ft2 ](7.5 ft) = 18.04ft3
and a specific volume of v1 = vg (14.696 psi) = 26.816 ft3 /lbm the mass is
M1 = 18.04ft3 /26.816 ft3 /lbm = 0.8408 lbm
Using the value for the work done calculated in Example 1 produces
w = W/M = 18900 ft lbf/[(778.2 ft lbf/Btu)(.8408lbm)] = 28.9 Btu/lbm
6.2.2 Watt’s Improvements

During the eighteenth century, many Newcomen engines were constructed and sold
and in the process, many improvements were made that enhanced their performance.
However, in 1776 James Watt introduced the first of his inventions that transformed
Newcomen’s atmospheric engine into the steam engine that supplied the power needs
of the Industrial Revolution and beyond. We will study the most important of these
improvements in detail.
The External Condenser
The cylinder in the Newcomen engine had two functions; in the language we learned
in Chap. 5:
• production of power, Ẇ S
• heating the low-temperature reservoir, Q̇ L/S
We know from our previous study that combining these functions in the same process
is not as efficient as using a piece of a Carnot cycle (recall Fig. 5.9). Watt sensed
this intuitively (he had no knowledge of the science of thermodynamics, which
would not be developed for another 50 years), and so designed an engine with an
external condenser. Thus, he split the cylinder into two elements, each with a separate
function:
• a prime mover to produce power without heat transfer
• a condenser to transfer heat without doing work
In order for the steam to exhaust from the prime mover into the condenser, the latter
had to be maintained below atmospheric pressure, and Watt provided a pump for this
purpose. Figure 6.4 is a schematic diagram of the arrangement of the four component
devices of Watt’s steam power plant.
Fig. 6.4 Flow diagram of a Watt engine showing the various work and heat transfer interactions.
The subscripts H and L are the hot and cold reservoirs, respectively, while I and O stand for the
input and output devices
Using Eq. (6.1) in the expression for w produces
w = W/M1 = pmep v1 [VD /(Vc + VD )] (6.5)
which shows that a given prime mover is more efficient (has a larger w) when it
has a higher mean effective pressure, pmep , and a larger specific volume, v1 . The
external condenser increases w on both these counts. Because the cylinder walls are
not cooled by water spray, indeed Watt tried to thermally insulate them from the
environment, v1 , is the same here as the boiler exit, namely, saturated vapor, x1 = 1.
Moreover, since the condenser can be maintained at a lower temperature than the
average Newcomen cylinder temperature, pmep is greater as well. These points are
quantified in the following example.
Example 6.3 A steam engine with the same dimensions as the Newcomen
engine of the previous example has its condenser maintained at 120 ◦ F by
the cooling water and a vacuum pump. What load can be raised by the prime
mover under these circumstances, how much work is done, and what is the
specific work done by the steam?
Solution.
Under these circumstances, the condenser pressure is 1.69 psi so that the mean
effective pressure is 13 psi. The load that can be raised is thus
Fe = (13 lbf/in2 )(346.36 in2 ) = 4160 lbf
which is 1.5 times the Newcomen value.

The work done by the steam is then
W = (4160 lbf)(7 ft) = 29100ft lbf
which is also 1.5 times the Newcomen value.

The specific work done by the steam is from Eq. (6.5)
w = (13 · 144 lbf/ft2 )(26.816 ft3 / lbm)(7/7.5)/(778.2 ft lbf/Btu) = 60.2 Btu/lbm
This is also a factor of 1.5 larger than the Newcomen value, which comes
from increasing the mean effective pressure.
Comparing this example with 6.2 shows that the engine with an external condenser
burns only 67% as much coal as the Newcomen engine.
The Expansive Power of Steam
Watt achieved another tremendous improvement in efficiency by recognizing that the
atmospheric steam would drive the piston itself against the condenser back pressure
so it was a waste to fill the entire cylinder. Consequently, he arranged to cut off the
steam flow at a fraction of the total cylinder volume. After this point, the expansion
of the steam in the cylinder followed an isentropic law (Watt made every effort to
reduce heat transfer from the cylinder) to the maximum stroke, then the exhaust valve
opened and the pressure dropped to the condenser level at constant volume, as shown
in Fig. 6.5. You can see in the figure that the total amount of work done is less than
the same size engine without steam cutoff. However the mass of steam required to
fill the cylinder under this condition, Vco /v1 , is also smaller, and the ratio produces
a larger w.
Fig. 6.5 Pressure–volume diagram with steam flow cutoff at Vco and a subsequent polytropic
expansion to the condenser pressure
Example 6.4 The engine of the previous example is cut off when the piston is
1.5 ft from the head of the cylinder, L co = 1.5 ft, and the steam subsequently
expands isentropically to the maximum volume. Calculate the total work done
and the specific work done by the steam per cycle of operation.
Solution.
For this configuration, the net work done by the steam is
W = W41 + W12s + W23
in which the first and last components are easily obtained from the accompa-
nying figure. The second component is the work done in an isentropic process
for which we need to know the value of the final pressure. From the figure,
this is
p2s = (1.5/7.5)1.29 · 14.7 psi
which has the value 1.844 psi. The work done by the steam is then
⎧
⎪
⎪ [ pb (L co − L c ) + ( p2s L T − pb L co )/(1 − k) − pc L D ]A p
⎨
{[14.7 · 1 − (1.844 · 7.5 − 14.7 · 1.5)/0.29−
W =
⎪
⎪ 1.7 · 7] ft lbf/in2 }(346.36 in2 )
⎩
10790 ft lbf
This is about a third of the previous value (see Example 6.3). However, now
the mass of steam expanded is
M1 = Vco /v1 = (18.04 ft3 )(1.5/7.5)/(26.816 ft3 /lbm)

which is 0.134 lbm and is only a fifth of the previous value. Consequently,
w = (10790 ft lbf)/[(0.134 lbm)(778.2 ft lbf/Btu) = 103.4 Btu/lbm
This is 1.7 times greater than the engine without cutoff and 3.6 times greater
than the original Newcomen engine.
The result of this example shows that an engine with these improvements would
burn less than 30% of the amount of coal used by a Newcomen engine that produced
the same amount of power.
Double Acting Rotating Machine
In another important development, Watt made the piston double acting. That is he
arranged for a power stroke in both directions of motion by means of pairs of inlet and
exhaust valves that were opened and closed by the motion of the piston. A schematic
of this is shown in Fig. 6.4. Not only did this double the work done in a cycle, but it
also removed the atmosphere from one side of the piston and made the device truly
a steam engine.
At the same time, he converted the output motion from reciprocating to rotating
by means of a mechanical linkage. One consequence of this was that the engine speed
was no longer primarily determined by the external load (refer to Fig. 6.3), but by
the rate of steam flow, through the two equations
{Ẇ } = W N = w{ Ṁ} (6.6)
These give the engine speed, using the definition of specific work done, as
N = { Ṁ}/M1 (6.7)
so a throttle in the steam supply line to the prime mover could be used to adjust its
speed by varying the mass rate of flow.
6.2.3 The Rankine Cycle
The science of Thermodynamics was developed in the attempt to analyze and under-
stand the operation of the steam engine. The discovery of the energy balance equation
in 1851 finally made this analysis possible; it was first performed by the Scottish engi-
neer, William Macquorn Rankine, after whom it is named, by applying it to each
of the four devices shown in Fig. 6.4. In this way the usual quantities, W , {Ẇ } and
w as well as ηT = {Ẇ }/{ Q̇} can be determined as functions of the boiler and con-
denser pressures, and the mass rate of flow of steam through the system. Note that the
specific fuel consumption, Eq. (6.4), is expressible in terms of the thermal efficiency
{ Ṁ f } 1 { Ṁ f } 1
sfc = = (6.8)
{ Ṁ} w { Q̇} ηT
Fig. 6.6 An ideal Rankine cycle, composed of four ideal devices (each with its energy equation),
is shown on the left. The state points on the T , s diagram on the right correspond to the locations
shown on the left. Two parameters specify the cycle, the boiler pressure, pb , and the condenser
pressure, pc . The net specific work done in a cycle is the area enclosed and shown shaded
The Rankine cycle is a thermodynamic analysis and so depends on the state changes
underwent by the steam as it flows from one device to the other rather than the
mechanical nature of the devices themselves. Therefore, the results of studying this
cycle are the same whether the prime mover is a piston moving in a cylinder or a
more modern steam turbine.
The Ideal Rankine Cycle

The basis of Rankine’s analysis is a plot, on a T, s diagram, of the changes in state
of an element of steam as it moves around the circuit depicted on the left in Fig. 6.6.
This plot, for an ideal Rankine cycle, is shown on the right in the figure. Note that
the cycle is completely determined by the boiler and condenser pressures. The inlet
to the prime mover is defined as saturated vapor at the boiler pressure, the prime
mover expands the steam isentropically to the condenser pressure, the condenser
transfers heat at the condenser pressure until the steam is saturated liquid, the pump
raises the liquid pressure isentropically to the boiler pressure, and the boiler transfers
heat at the boiler pressure until the steam is saturated vapor. The following example
illustrates these calculations. In the example, x2 is calculated from Eq. (5.25) and h 2
from Eq. (4.43) using values in the table of water.
Example 6.5 Calculate the net specific work done and the thermal efficiency
of an ideal Rankine cycle operating between a boiler pressure of 14.7 psi and
a condenser pressure of 1.6927 psi (saturation temperature of 120 ◦ F).
Solution.
State 1 p1 = 14.7 psi x1 = 1
h1 (p1 , x1 ) = 1150.5 Btu
lbm s1 (p1 , x1 ) = Btu
1.7569 lbm R
State 2 Tf 2 = 120◦ F s2 = s1
1.7569−0.1646
x2 (s2 , Tf 2 ) = 1.933−0.1646 = 90.0% h2 (x2 , Tf 2 ) = [87.97 + x2 (1113.6 − 87.97)] Btu
lbm
h2 = 1011.0 Btu lbm wt = h1 − h2 = 139.50 Btu
lbm p2 (Tf 2 ) = 1.6927 psi
State 3 p3 = p2 x3 = 0
h3 (p3 , x3 ) = 87.97 Btu
lbm
Btu
s3 (p3 , x3 ) = 0.1646 lbm R ql = h3 − h2 = −923.03 Btu
lbm
State 4 p4 = p1 s4 = s3
h4 (s4 , p4 ) = 88.009 Btu
lbm wp = h4 − h3 = 0.04 Btu
lbm qh = h1 − h4 = 1062.49 Btu
lbm
Cycle Quantities w = wt − wp = 139.46 Btu
lbm ηT = w/qh = 13.1%
Note that the net heat absorbed q = qh + ql = 139.46 Btu/lbm is equal to w as it must
be for a closed cycle. Also, the pump work is much smaller than the turbine work and
therefore to a rough approximation, it is zero and h 4 ∼ h 3 . A better approximation
is provided by the result of Example 5.1, w p ∼ v3 ( p4 − p3 ). Using this, there is no
need to calculate h 4 and then determine w p = h 4 − h 3 . When finding h 4 requires
an interpolation, we can use w p ∼ v3 ( p4 − p3 ) to determine w p , and then use h 4 =
h 3 + w p to find h 4 .
The prime mover, condenser, and pump of the ideal Rankine cycle are perfect
devices in the sense that no entropy is generated in any of them. Thus, for example, the
work done by the prime mover in this cycle is the equilibrium work done between the
initial and final states. This fact allows us to evaluate the thermodynamic efficiency
of the Watt engine of Example 6.4 (since the conditions are the same in the two
examples) by calculating the ratio w/w∗ ; this is 103.4/139.46 = 0.741. Although this
is not too bad, present-day prime movers (steam and gas turbines) have efficiencies
in the range of 85–90 percent.
The ideal Rankine cycle differs from a Carnot cycle between the boiler and con-
denser temperatures only because the boiler process is constant pressure rather than
constant temperature. This is done for practical reasons and the penalty is not large.
Note that the Carnot efficiency for the boiler and condenser temperatures of example
6.5 is ηT∗ = 1 − (120 + 460)/(212 + 460) = 13.7%, so the corresponding thermo-
dynamic efficiency of the Rankine cycle of this example is η = ηT /ηT∗ = 0.962.
The hot reservoir temperature is actually some 2000K corresponding to the
combustion temperature of the fuel–air mixture. Therefore, considering the low-
temperature reservoir to be around 300K, the theoretical maximum efficiency of
such a power plant is the Carnot value 1 − 300/2000 = 0.85 so the corresponding
thermodynamic efficiency of the cycle in Example 6.5 is 0.155. Practical consider-

ations immediately enter to modify this limit for a steam power plant, because steel
piping has no structural integrity at such high temperatures. In this way, the practical
maximum hot reservoir temperature is around 800K (500 ◦ C) with a corresponding
Carnot efficiency of 0.625. This is still much higher than the 0.132 we just calculated
for the cycle of example 6.5, the corresponding thermodynamic efficiency is 0.211.
and our subsequent discussion is intended to describe the various methods that are
in use to raise these values.
The simplest ways to raise the efficiency of the ideal Rankine cycle are
• increase the boiler pressure,
• decrease the condenser pressure.
You can see this by forming a Carnot efficiency based on the boiler and condenser
temperatures, ηT ∼ 1 − Tc ( pc )/Tb ( pb ), as an approximation to the actual cycle effi-
ciency. The condenser pressure is the saturation pressure at the condenser temper-
ature. This temperature is set by the temperature of the cooling water and the heat
transfer characteristics of the condenser. It is lower in the winter than the summer,
but in any event cannot be varied by the power plant designer. The boiler pressure
can be far above atmospheric pressure; however, as the boiler pressure increases, the
quality of steam at the turbine outlet decreases. Since qualities less than about 90%
will cause turbine blade failure due to erosion, the boiler pressure is severely limited
on this account; the quality is already 90% in Example 6.5 and any increase in boiler
pressure will decrease it.
Superheat
By continuing to transfer heat to the saturated vapor, we can raise its temperature
along the constant pressure, pb , curve so that the boiler output state is in the vapor
region. This is called superheating the steam. By doing this, we can increase the
steam quality at the turbine outlet. Together with an increase of the boiler pressure,
superheat provides a way to substantially increase the cycle efficiency. The following
example illustrates the effect of superheat together with an increase of boiler pressure
on the Rankine cycle.
of a Rankine cycle with superheat. The boiler pressure is 100 psi and the
temperature at the turbine inlet is 600 ◦ F; the condenser pressure is 1.6927 psi
(120 ◦ F) as before.
Solution.
State 1 p1 = 100 psi Tf 1 = 600◦ F
h1 (p1 , Tf 1 ) = 1329.6 Btu
lbm s1 (p1 , Tf 1 ) = Btu
1.7586 lbm R
State 2 Tf 2 = 120◦ F s2 = s1
x2 (s2 , Tf 2 ) = 1.7586−0.1646
1.933−0.1646 = 90.1% h2 (x2 , Tf 2 ) = [87.97 + x2 (1113.6 − 87.97)] Btu
lbm
h2 = 1012.0 Btu lbm wt = h1 − h2 = 317.6 Btu
lbm p2 (Tf 2 ) = 1.6927 psi
State 3 p3 = p2 x3 = 0
h3 = (p3 , x3 ) = 87.97 Btu
lbm s3 (p3 , x3 ) = Btu
0.1646 lbm R ql = h3 − h2 = −924 Btu
lbm
State 4 p4 = p1 s4 = s3
h4 (s4 , p4 ) = 88.26 Btu
lbm wp = h4 − h3 = 0.30 Btu
lbm qh = h1 − h4 = 1241.3 Btu
lbm
Cycle Quantities w = wt − wp = 317.3 Btu
lbm ηT = w/qh = 25.6%
Comparing the results of this example to those of 6.5 shows that ηT has increased
while x2 has remained at about 90% (T f 1 was chosen so that this would be true).
Each of these quantities continually increase as T f 1 increases, but as we mentioned
before this quantity is limited by the mechanical strength of the pipes that carry
the steam. For the conditions of Example 6.5 but with a boiler pressure of 1540
psi (corresponding to a boiler saturation temperature of 600 ◦ F), the cycle thermal
efficiency is much higher (37%) than that of Example 6.6; however, the quality is far
below 90% (only 66%). Using superheat in this way permits an increase in thermal
efficiency to 33.2%, corresponding to a boiler pressure of 300 psi and a limiting
turbine inlet temperature of about 900 ◦ F while maintaining a quality of 90% at the
turbine outlet. Further improvements of efficiency can be obtained either by raising
the boiler pressure together with another mechanism for increasing the turbine exit
quality or by using an adaptation here of Watt’s idea of using the expansive power
of the steam.
Reheat
In this concept, the turbine is split in two (or more) parts and between them the
steam is returned to the boiler and again heated so that its temperature is restored to a
suitably high value. We call this method reheating. A flow diagram and corresponding
T, s diagram are drawn in Fig. 6.7. Note that in this cycle the turbine work is wt =
whp + wlp while the heat transfer from the source is the sum of that of the main
boiler and the reheater, qh = qb + qr . The following example illustrates the increased
efficiency that is possible with reheat.
Fig. 6.7 A Rankine cycle with reheat is shown on the left. The state points on the T , s diagram on
the right correspond to the locations shown on the left
of a Rankine cycle with superheat and reheat. The boiler pressure is 800 psi
and the temperature at the high-pressure turbine inlet is 600 ◦ F. This turbine
has an exit pressure of 300 psi and the steam in this state is reheated in the
boiler to 900 ◦ F. This steam then is expanded in the low- pressure turbine to
the condenser pressure, which is 1.6927 psi (condenser temperature 120 ◦ F)
as before.
Solution.
State 1 p1 = 800 psi Tf 1 := 600◦ F
h1 (p1 , Tf 1 ) = 1271.1 Btu
lbm
Btu
s1 (p1 , Tf 1 ) = 1.4869 lbm R
State 2 p2 = 300 psi s2 = s1
1.4869−0.5882
x2 (s2 , p2 ) =1.5105−0.5882 = 97.4% h2 (x2 , p2 ) = [393.99 + x2 (1202.9 − 393.99)] Btu
lbm
h2 = 1182.2 Btu
lbm
Btu
wht = h1 − h2 = 88.9 lbm
State 3 p3 = p2 Tf 3 = 900◦ F
h3 (p3 , Tf 3 ) = 1473.6 Btu
lbm
Btu
s3 (p3 , Tf 3 ) = 1.7591 lbm R qr = h3 − h2 = 291.4 Btu
lbm
State 4 p4 = 1.6927 psi s4 = s3
x4 (s4 , p4 ) = 90.1% h4 (x4 , p4 ) = 1012.3 Btu
lbm wlt = h3 − h4 = 461.3 Btu
lbm
State 5 p5 = p4 x5 = 0
3
h5 (x5 , p5 ) = 87.97 Btu
lbm
Btu
s5 (x5 , p5 ) = 0.1646 lbm R
ft
v1 (x5 , p5 ) = 0.16204 lbm
State 6 p6 = p1 s6 = s5
wp = v5 (p6 − p5 ) = 2.39 Btu
lbm h6 = h1 + wp = 90.36 Btu
lbm qb = h1 − h6 = 1180.7 Btu
lbm
Cycle Quantities qh = qb + qr = 1472.1 Btu
lbm wt = wht + wlt = 550.2 Btu
lbm
w = wt − wp = 547.8 Btu
lbm ηT = w/qh = 37.2%
From these results, we see that the cycle thermal efficiency has been increased to
37.2% by means of using reheat together with superheat and higher boiler pressure,
while maintaining both turbine outlet qualities above 90%. Also, we used here the
approximate formula for w p , in order to avoid having to interpolate to find h 6 .
Regeneration
As with reheat, regeneration entails splitting the turbine in two parts, but instead of
reheating the steam, part of it is diverted from the low-pressure turbine and instead is
used to preheat the output water from the condenser prior to sending it to the boiler.
In effect, this steam is used for heating instead of producing work and the increased
efficiency results from its more effective utilization in this capacity.
The heating of the boiler feedwater (condenser output) can be done either by
mixing in an open feedwater heater (see Sect. 4.4.2, Mixing Chambers and Heat
Exchangers) as shown in Fig. 6.8, or by heat transfer in a closed feedwater heater
(see Sect. 4.4.2, Mixing Chambers and Heat Exchangers), shown in Fig. 6.9. The
open feedwater heater requires two pumps as can be seen in the figure while the
closed feedwater heater needs only one pump and a trap, which is another name for
a throttle (see Sect. 4.4.2, Duct Flow).
In either case at the intermediate pressure, which we call the extraction pressure
a fraction y = Ṁe / Ṁ, where Ṁ is the mass flow rate of water through the boiler,
is extracted from the flow prior to entering the low-pressure turbine. In this event,
the power developed by the low-pressure turbine is Ẇlt = ( Ṁ − Ṁe )h, with h
the enthalpy difference across this turbine. The specific work done by this device
obtained by dividing by the mass rate of flow is then wlt = (1 − y)h.
Given the extraction pressure and the states entering and leaving the feedwater
heater, the mass and energy balances for this device specify the extraction fraction.
For example, in the system depicted in Fig. 6.8 the energy balance for the open
feedwater heater is
h 6 Ṁ = h 2 Ṁe + h 5 ( Ṁ − Ṁe )
and the mass balance gives Ṁ = Ṁe + ( Ṁ − Ṁe ). Dividing by Ṁ and solving for
y produces the result
h6 − h5
y=
h2 − h5
For the system with the closed feedwater heater shown in Fig. 6.9, the energy and
mass balances yield
h6 − h5
y=
h2 − h7
for the fraction extracted. Note that the output state of the extracted fluid, state 6
in Fig. 6.8 or state 7 in Fig. 6.9 is specified to be saturated liquid at the extraction
pressure. In addition, the temperature at state 6 in Fig. 6.9 must be specified as well.
Fig. 6.8 A Rankine cycle with regeneration using an open feedwater heater is shown on the left.
The state points on the T , s diagram on the right correspond to the locations shown on the left. The
extraction fraction, denoted by y, is determined by requiring state 6 to be saturated liquid
Fig. 6.9 A Rankine cycle with regeneration using a closed feedwater heater is shown on the left.
The state points on the T , s diagram on the right correspond to the locations shown on the left.
The extraction fraction, denoted by y, is specified by giving the temperature at state 6 and requiring
state 7 to be saturated liquid
These calculations are illustrated in the following example, which adds a feedwater
heater to the superheat cycle of example 6.6. In this example, the quantity f 2 is an
interpolation variable on entropy and h 2 is the corresponding interpolated value for
enthalpy from values in the table for water.
of a Rankine cycle with regeneration as shown in Fig. 6.8. The boiler pressure is
100 psi and the temperature at the turbine inlet is 600 ◦ F; the extraction pressure
is 20 psi, the condenser pressure is 1.6927 psi, and the system includes an open
feedwater heater.
Solution.
State 1 p1 = 100 psi Tf 1 = 600deg F
h1 = (p1 , Tf 1 ) = 1329.6 Btu
lbm
Btu
s1 (p1 , Tf 1 ) = 1.7586 lbm R
State 2 p2 = 20 psi s2 = s1
1.7586−1.7406
f2 = 1.7676−1.7406 = 0.6667 h2 (s2 , p2 ) = [(1162.2 + f2 (1181.7 − 1162.2)] Btu
lbm
h2 = 1175.2 Btu
lbm wht = h1 − h2 = 154.4 Btu
lbm
State 3 p3 = 1.6927 psi s3 = s2
1.7586−0.1646
x3 (p3 , s3 ) =1.9339−0.1646 = 90.1% h3 (p3 , s3 ) = [87.97 + x3 (1113.6 − 87.97)] Btu
lbm
h3 = 1011.6 Btu
lbm wlt = h2 − h3 = 163.6 Btu
lbm
State 4 p4 = p3 x4 = 0
h4 (x4 , p4 ) = 87.97 Btu
lbm
Btu
s4 (x4 , p4 ) = 0.1646 lbm R
State 5 p5 = p2 s5 = s4
h5 (s5 , p5 ) = 88.02 Btu
lbm wlp := h5 − h4 = 0.05 Btu
lbm
State 6 p6 = p5 x6 = 0
h6 −h5
h6 (x6 , p6 ) = 196.27 Btu
lbm v6 (x6 , p6 ) = Btu
0.01683 lbm R y= h2 −h5 = 9.957%
State 7 p7 = p1 s7 = s6
whp = v6 Δp = 0.25 Btu
lbm h7 = h6 + whp = 196.52 Btu
lbm qh := h1 − h7 = 1133.1 Btu
lbm
Cycle Quantities wt := wht + (1 − y)wlt = 301.7 Btu
lbm
wp = whp + (1 − y)wlp =).3 Btu
lbm w = wt − wp = 301.4 Btu
lbm
ηT = w/qh = 26.6%
Comparing the results of this example to those of 6.5 shows that ηT has increased
here by 1.1% as a result of regeneration.
Closed feedwater heaters, as shown in Fig. 6.9, are also used in modern power
plants. Indeed, modern steam power cycles use both reheat and superheat, as well
as more turbine stages with higher pressures in order to achieve higher thermal
efficiencies.
Turbine and Pump Losses
The largest losses in efficiency are caused by inefficiencies in the turbines. Pump
losses are not as significant because, as we have seen before, the pump work itself is
a small fraction of the turbine work. For an expanded discussion of these points refer
to Sect. 5.5.5, Open Systems. In particular, the results given here for the adiabatic
efficiency of turbines and compressors are obtained in Eqs. (5.61) and (5.62) and
illustrated in Fig. 5.10.
Turbine efficiency is measured by comparing the actual adiabatic turbine work
wt = h in − h out (6.9)
with that of an isentropic turbine operating between the same two pressures,
wt∗ = h in − h ∗out (6.10)

where h ∗out = h(sin , pout ). The efficiency is ηt = wt /wt∗ . Thus given ηt , the input
state, and output pressure, we first calculate wt∗ , from Eq. (6.10), then wt = wt∗ ηt ,
and finally, from Eq. (6.9), h out = h in − wt . On the other hand, given the input and
output states, we calculate wt from Eq. (6.9) and wt∗ from Eq. (6.10), then finally
the turbine efficiency from its definition. In the next example, we illustrate these
calculations by modifying Example 6.5 to have a turbine efficiency less than one.
of an ideal Rankine cycle operating between a boiler pressure of 14.7 psi and
a condenser pressure of 1.6927 psi (saturation temperature of 120 ◦ F) when
the turbine has an efficiency of 87%
Solution.
State 1 p1 = 14.7 psi x1 = 1
h1 (p1 , x1 ) = 1150.5 Btu
lbm s1 (p1 , x1 ) = Btu
1.7569 lbm R
State 2s Tc2s = 120◦ F s2s = s1
x2s (s2s , Tc2s ) = 1.7569−0.1646
1.933−0.1646 = 90.0% h2s (x2s Tc2s ) = [87.97 + x2 (1113.6−
87.97)] lbm = 1011.0 Btu
Btu
lbm wt∗ = h1 − h2s = 139.50 Btu ∗ Btu
lbm wt := wt ηt = 121.4 lbm
State 2 h2 = h1 − wt = 1029.1 Btu
lbm p2 = p2s = 1.6927 psi
1029.1−87.97 Btu
x2 (h2 , p2 ) = 1113.6−87.97 = 91.8% s2 (x2 , p2 ) = 1.7882 lbm R
State 3 p3 = p2 x3 = 0
h3 (p3 , x3 ) = 87.97 Btu
lbm s3 (p3 , x3 ) = Btu
0.1646 lbm R ql := h3 − h2 = −941.1 Btu
lbm
State 4 p4 = p1 s4 = s3
h4 (s4 , p4 ) = 88.01 Btu
lbm wp := h4 − h3 = 0.04 Btulbm qh = h1 − h4 = 1062.5 Btu
lbm
Cycle Quantities w = wt − wp = 121.4 Btu lbm ηT = w/qh = 11.4%
Btu
Note that s2 − s2s = 0.0313 lbm R
This result for thermal efficiency is just 87% of the value in Example 6.5 because
only the turbine work has its value changed. The pump work can be treated in an
analogous way; namely, the actual adiabatic pump work is
w p = h out − h in
while the ideal isentropic value is, for the same pressures,
w ∗p = h ∗out − h in
Here h ∗out = h(sin , pout ) as before, but now the pump efficiency2 is η p = w ∗p /w p .
Thus given η p , the input state, and output pressure, we first calculate w∗p , then w p =
w ∗p /η p and finally h out = h in + w p .
2 Note the different definitions for the turbine and pump efficiencies; this happens because the actual
pump work is greater than the ideal while the actual turbine work is less.
6.3 Air Standard Power Cycles 251
6.3 Air Standard Power Cycles

An alternative to using the products of combustion to boil water and subsequently to
use the steam to create mechanical energy, is to use them directly. This idea was devel-
oped about a hundred years after Watt had worked on the steam engine. Although the
initial attempts to construct an internal (as opposed to external) combustion engine
were implemented by a piston moving in a cylinder and connected to a slider crank
mechanism in which the thermodynamic cycle was completed in one crank revo-
lution, called a two-stroke engine, the first commercially successful engines were
based on a four-stroke process in which one thermodynamic cycle corresponded to
two crank revolutions. The idea for this was patented by Alphonse Beau de Rochas,
a French Engineer, in 1862. The device, shown in Fig. 6.10, operates as follows:
• Starting at top dead center (TDC) and with the intake valve open and exhaust
valve closed, the piston moves downward to bottom dead center (BDC) with a
180-degree rotation of the crankshaft, while filling the volume with an air–fuel
mixture
• With both valves closed, the crank rotates another 180 degrees and the piston
moves back up to TDC, compressing the substance in the cylinder
• The mixture is ignited and burns so quickly that in this time the piston does
not move; the high-pressure products of combustion then expand to the volume
corresponding to BDC while the crankshaft rotates another 180 degrees
• The exhaust valve opens and as the piston moves back up to TDC with a cor-
responding crankshaft rotation of 180 degrees, the burned fuel–air mixture is
expelled
• On the next downstroke, the intake valve is open and the process is repeated.
Just as with the steam engine, the geometric quantities that define the internal
combustion engine are the clearance volume Vc , between TDC and the top of the
cylinder, and the displacement (displacement volume) VD = Ac L = π(D/2)2 L, the
volume swept by the piston during its motion, which is just the volume between
Fig. 6.10 Schematic of a

four-stroke piston engine
showing the top and bottom
dead center locations of the
piston head
TDC and BDC. In terms of these, the compression ratio r is important in describing
engine performance. This equation can be used to determine the clearance volume
in terms of the displacement and compression ratio
Vc + VD VD
r= −→ Vc = (6.11)
Vc r −1
The mean effective pressure, Eq. (6.1), in the present situation is written as
mep = W/VD = w/v D (6.12)
The mep is a measure of the efficacy of the engine . A higher mep indicates an engine
that produces a higher power than another of the same size; it has a higher power
density.
6.3.1 The Otto Cycle

The first four-stroke engine was built in 1876 by Nikolaus A. Otto. This was the
first practical alternative to the steam engine as a power source. Moreover, its higher
power density (power per unit mass) made it more suitable for mobile applications
(automobile, airplane).
The cycle shown in Fig. 6.11 is named in Otto’s honor. In this idealized represen-
tation, called an Air Standard Cycle, the isentropic process from state 1 to 2 models
the compression stroke and that from 3 to 4 the expansion stroke. Combustion occurs
so quickly that the piston has no time to move and so is an isochoric process. Finally,
the purging of the burned mixture and recharging of the cylinder with a fresh one is
replaced by heat rejection at constant volume.
Assuming that the working substance is air as an ideal gas, and given State 1,
the compression ratio, and the top cycle temperature T3 , all quantities relating to
the thermodynamic cycle can be calculated, including the specific work done. If in
Fig. 6.11 The air standard Otto cycle consists of four processes of a closed system as an approxi-
mation to the actual engine performance. Both compression and expansion processes are assumed
isentropic while the heat addition and rejection processes are assumed constant volume
addition, one knows the displacement, the total work done is known as well; this is the
engine torque. Finally, if the engine rotational speed is known, the power developed
is also known. The following example illustrates these points. In the example f is
an interpolation variable, once on T and once on vr , with u and vr corresponding
interpolated values, all from values in the table for air.
Example 6.10 A 6 cylinder engine takes in an air–fuel mixture at 14.7 psi and
50 ◦ F and compresses it by a factor of 10. The top temperature reached during
combustion is 3500 ◦ F. Calculate the specific work done in one cycle and the
specific heat transfer. If the engine bore is 3.5 in and the stroke is 3.25 in,
calculate the work done and mep. When the engine is running at 3000 rpm,
calculate the power developed.
Solution.
r = 10 D = 3.5 in L = 3.25 in
VD
VD = π( D 2 3
2 ) L = 31.269 in Vc = r−1 = 3.474 in3
State 1 p1 = 14.7 psi T1 = 509.67R
RG T1 ft3 V1
V1 = Vc + VD = 34.743 in3 v1 = p1 = 12.843 lbm M1 := v1 = 0.001565lbm
u1 (T1 ) = −22.901 Btu
lbm vr1 (T1 ) = 175.53
v1 vr1
State 2 v2 = r vr2 = r = 17.553
vr2 −16.704
f2(vr2 ) = 17.659−16.704 = 0.8890 u2 (f2 ) = 107.51 Btu
lbm
wc = u1 − u2 = −130.41 Btu
lbm
State 3 Tf 3 = 3500◦ F v3 = v2
vr3 (Tf 3 ) = 0.46705 u3 (Tf 3 ) = 700.97 Btu
lbm qh = u3 − u2 = 593.46 Btu
lbm
State 4 v4 = v1 vr4 = vr3 r = 4.6705
f4 (vr4 ) = 0.809 u4 (f4 ) = 250.78 Btu
lbm we = u3 − u4 = 450.19 Btu
lbm
Cycle Quantities
Btu w rw
w = we + wc = 320 lbm ηT = qh = 54% mep = v1 (r−1) = 150 psi
1
W = 6wM1 = 2340 ft lbf N= 1500 min P = W N = 106 hp
In making the isentropic calculations here we used vr because r is a volume ratio

(see Sect. 5.3.2, Ideal Gases). We may notice that the thermal efficiency is much
higher than what can be achieved in a Rankine cycle. This occurs because in the Otto
cycle the top cylinder temperature is instantaneously much higher than what can be
maintained continuously in the boiler piping of a Rankine cycle. This is reflected in
the fact that using T3 and T1 of Example 6.10 as the hot and cold reservoir temperatures
in a Carnot cycle produces a theoretical maximum thermal efficiency for this case,
ηT∗ = 87.1%.
If we assume that the air is a perfect gas, that is, one with constant specific
heats, its mechanical and energetic equations of state are simple enough to allow
us to easily write compact expressions for specific work, heat transfer, and thermal
efficiency. Thus from the equation of state u = u 0 + cv (T − T0 ) (see Sect. 4.3.2,
Perfect Gases), we find for the thermal efficiency

cv (T3 − T4 ) − cv (T2 − T1 ) T3 1 − T4
T3
− T2 1 − T1
T2
ηT = =
cv (T3 − T2 ) T3 − T2
Since the two processes 1 to 2 and 3 to 4 are isentropic, the ratios T4 /T3 and T1 /T2
are the same and are equal to r 1−k (see Sect. 5.3.2, Perfect Gases), where k is the
ratio of the specific heats. As a consequence, on canceling like terms in the numerator
and denominator
1
ηT = 1 − k−1 (6.13)
r
In this way we see that the efficiency depends only on the compression ratio. Using
the value k = 1.4 and a compression ratio of 10, we find from Eq. (6.13) that ηT =
60.2%, which is about 14% higher than the result in Example 6.10. This difference
is caused by the fact that air is not a perfect gas, its specific heats change with
increasing temperature. Still an equation like Eq. (6.13) is useful because we can see
the important parameters that make it vary, as well as calculate its approximate value
rather easily.
Although Eq. (6.13) tells us that we want to make the compression ratio as large
as possible, we reach a limit in the Otto cycle when the temperature T2 is so high that
it causes the air–fuel mixture to ignite on its own (called pre-ignition or knocking).
An alternative cycle that avoids this problem was developed by Rudolf Diesel in the
1890s, and a successful engine was produced in 1897.
6.3.2 The Diesel Cycle
Diesel’s idea was to compress air alone to a temperature high enough to spontaneously
ignite the fuel and then to spray in the fuel at a measured rate. In this way, the
combustion process occurred at constant pressure rather than at constant volume as
in the Otto cycle. The resulting cycle, shown in Fig. 6.12, depends on the compression
ratio r = V1 /V2 and the cutoff ratio rc = V3 /V2 . The following example illustrates
a typical calculation.
Fig. 6.12 The air standard Diesel cycle consists of four processes of a closed system as an approx-
imation to the actual engine performance. Both compression and expansion processes are assumed
isentropic, and the heat rejection is assumed constant volume as in the Otto cycle, but now the heat
addition is assumed constant pressure so qh = h 3 − h 2
Example 6.11 A 6-cylinder diesel engine takes in air at 14.7 psi and 50 deg_F,
compresses it by a factor of 17, and operates with a cutoff ratio of 3. Calculate
the specific work done in one cycle and the specific heat transfer. If the engine
bore is 3.5 in and the stroke is 3.25 in, calculate the work done and mep. When
the engine is running at 3000 rpm, calculate the power developed.
Solution.
D = 3.5 in L = 3.25 in r = 17
VD
VD = π( D
2)
2 = 31.269 in3 Vc = r−1 = 1.954 in3 rc = 3
State 1 p1 = 14.7 psi T1 = 509.67R
3
RG T1
V1 = Vc + VD = 33.223 in 3
v1 = p1 = 12.846 ft M1 = V1
v1 = 0.001497lbm
lbm
u1 (T1 ) = −22.901 Btu
lbm vr1 (T1 ) = 175.53
v1 vr1
State 2 v2 = r vr2 = r = 10.325
f2 (vr2 ) = 0.3239 u2 (f2 ) = 157.18 Btu
lbm h2 (f2 ) = 259.56 Btu
lbm
wc = u1 − u2 = −180.08 Btu
lbm T2 (f2 ) = 1493.5R
State 3 p3 = p2 v3 = v2 rc
T3 = T2 rc = 4480.5R f3 (T3 ) = 0.4100 h3 (f3 ) = 1131.21 Btu
lbm
vr3 (f3 ) = 0.30508 u3 (f3 ) = 824.10 Btu
lbm qh = h3 − h2 = 871.66 Btu
lbm
wb = p3 v1 (rcr−1) = 204.75 Btu
lbm
State 4 v4 = v1 vr4 = vr3 rrc = 1.7288
f4 (vr4 ) = 0.7306 u4 (f4 ) = 406.50 Btu
lbm we = u3 − u4 = 417.60 Btu
lbm
Cycle Quantities
w = wb + we + wc = 442.3 Btu
lbm ηT = w
qh = 50.7% rw
mep= v1 (r−1) = 198 psi
1
W = 6wM1 = 3090 ft lbf Nc = 1500 min P = W Nc = 140 hp
As we did with the Otto cycle, we can obtain simple expressions for the cycle quan-
tities for the Diesel cycle by assuming that the air is a perfect gas. In this way we can
obtain k
1 rc − 1
ηT = 1 − k−1 (6.14)
r k(rc − 1)
for the thermal efficiency. Equation (6.14) gives 58% on using the values of Example
6.11 and is about 12% higher than the value calculated there. Again this difference
is a result of the perfect gas assumption of constant specific heats, c p and cv . The
quantity in square brackets in Eq. (6.14) is an increasing function of rc and is equal
to 1 in the limit rc → 1, so a Diesel (compression ignition) engine with the same
compression ratio as an Otto (spark ignition) engine has a lower thermal efficiency.
However, it can be operated at higher compression ratios with higher efficiencies
than are possible with the Otto cycle.
6.3.3 The Brayton Cycle
Another engine type that uses fuel burned in air directly is the gas turbine. We have
seen previously (Sect. 6.2.3) how steam turbines can be used in power producing
applications; gas turbines can be used similarly and we will study these implemen-
tations in this section.
Gas turbines are steady flow devices and their analysis, therefore, must be different
than the closed system, piston cylinder engines studied in this section. However, while
power plants based on them operate by continuously taking in air and fuel, burning
them together, and then exhausting them back to the surroundings (in other words as
an open system), we analyze them as an equivalent air standard (closed) system by
inserting a lossless, constant pressure, heat exchanger connecting the turbine output
to the compressor input. Such an idealized system is shown in Fig. 6.13 along with
its T , s diagram. This cycle, consisting of an isentropic compressor and isentropic
Fig. 6.13 The flow diagram of a gas turbine power plant and its air standard Brayton cycle
turbine, together with constant pressure heat exchangers one of which represents the
combustor, is called a Brayton cycle after George B. Brayton.3
In addition to the usual cycle quantities, we calculate for these cycles the back
work ratio
|wc |
bwr = (6.15)
wt
This quantity was not important in Rankine cycles because the pump work was always
small, but here the compressor work can be a significant fraction of the turbine work.
Indeed, early implementations of this type of power plant failed on account of this
factor.
If we are given the compressor inlet state, its pressure ratio and the turbine inlet
temperature, we can calculate all the intensive thermodynamic information related
to the cycle, because the turbine pressure ratio is the same as that of the compressor
Example 6.12 A gas turbine power plant operates with air of 14.7 psi, 50 ◦ F,
and 10 ft/s entering the compressor. The compressor pressure ratio is 10 and
the turbine inlet temperature is 2550 ◦ F. Calculate the net specific work, back
work ratio, heat transfer to the air, and thermal efficiency. In addition calculate
the mass rate of flow and the power developed, knowing that the inlet diameter
of the compressor is 2 ft.
Solution.
rp = 10 D1 = 2 ft A1 = π( D21 )2 = π ft2
State 1 p1 = 14.7 psi T1 = 509.67R
3 RG T1 3
v1 = 10 fts V̇1 = v1 A1 = 10π fts v1 = p1
ft
= 12.843 lbm
˙
Ṁ = Vv11 = 2.446 lbm
s h1 (T1 ) = 12.037 Btu
lbm pr1 (T1 ) = 1.4370
State 2 p2 = p1 rp pr2 = pr1 rp = 14.370
f2 (pr2 ) = 0.5550 h2 (f2 ) = 125.31 Btu
lbm wc = h1 − h2 = −113.28 Btu
lbm
State 3 p3 = p2 Tf 3 = 2550◦ F
h3 (Tf 3 ) = 688.17 Btu
lbm pr3 (Tf 3 ) = 1268.1 qh = h3 − h2 = 562.86 Btu
lbm
pr3
State 4 p4 = p1 pr4 = rp = 126.81
f4 (pr4 ) = 0.3290 h4 (f4 ) = 322.29 Btu
lbm wt = h3 − h4 = 365.88 Btu
lbm
Cycle Quantities
w = wt + wc = 252.61 Btu
lbm bwr= |w c|
wt = 31.0% ηT = w
qh = 44.9%
Ẇ = wṀ = 874 hp
3 Braytondid no work on gas turbines, but he had earlier built a two-stroke internal combustion
engine whose idealized cycle looked like the T -s diagram in Fig. 6.13. That is the reason this cycle
is named for him.
Here the efficiency is quite small and the back work ratio is large, because the work
required to compress the air is a significant fraction of the work done when the air
expands through the turbine.
As we saw with the other air standard cycles we have studied, the simple nature
of the perfect gas equations of state allow us to obtain simple expressions for the
important cycle quantities. Thus the thermal efficiency is
1
ηT = 1 − (k−1)/k
(6.16)
rp
where r p = p2 / p1 is the pressure ratio across the compressor. Using this with the
value 10 from the previous example gives 48.2%, which is 7% higher due to the
neglect of the variation of c p with temperature.
As with the steam turbine, there is a maximum turbine inlet temperature, imposed
by metallurgical requirements, of about the 3000 R used in our example. This means
that if we were to use the maximum pressure ratio across the compressor, the one
that produces that temperature from compression, without burning fuel, we would
obtain no net work. Since we would also obtain zero net work from not compressing,
or expanding, the air at all, it is clear from the first law of calculus that the net work
has a maximum at some intermediate pressure ratio. For a perfect gas

T4 T2 T1
w = c p (T3 − T4 ) − c p (T2 − T1 ) = c p T3 1 − − +
T3 T3 T3
(k−1)/k
and T2 /T1 = r p = T3 /T4 , we have
w T1 T2 T1
=1− − +
c p T3 T2 T3 T3
As we argued above, this is zero when T2 = T3 and when T2 = T1 . The maximum

to T2 is zero; namely,
value of this function occurs when the derivative with respect√
when T1 /T22 − 1/T3 = 0 or, in other words, when T2 max = T1 T3 , the geometric
mean of T1 and T3 . This corresponds to the pressure ratio

k/[2(k−1)]
T3
r p max = (6.17)
T1
and the maximum value of the specific net work

wmax = c p ( T3 − T1 )2 (6.18)
Using the data of the previous example in these equations gives a pressure ratio
of 22.4 and a maximum specific work of 250.2 Btu/lbm when T f 2 = 779 ◦ F.
Changing the pressure ratio by trial and error in Example 6.12 4 produces 33 and
4 This
was not done by looking up function values in the Air as an Ideal Gas Tables, but with a
computer analysis using calculated values of the properties.
267 Btu/lbm for the pressure ratio and net work done. In using Eqs. (6.17) and (6.18),
you must use absolute temperatures in order to arrive at the correct numerical value.
As before, these differences are caused by the variation of the specific heats of air
with temperature.
Reheat and Intercooling
The efficiency of the Brayton cycle can be increased by using the same ideas that
were used in the Rankine cycle, reheat and regeneration. In addition, because the
compressor work is significant in this cycle, it is worthwhile to reduce it as much as
possible. The intercooling concept does this and improves compressor performance
analogous to the way reheat improves turbine performance.
In the flow diagram shown in Fig. 6.14, the heat transfer from the hot reservoir
is qh = qh1 + qh2 , the turbine work is wt = wt1 + wt2 and the compressor work is
wc = wc1 + wc2 . Then the net work is still w = wt + wc and the thermal efficiency
is ηT = w/qh . The following example shows how these ideas are implemented in a
cycle (Fig. 6.15).
Fig. 6.14 The flow diagram of a gas turbine power plant with reheat and intercooling
Fig. 6.15 The T -s diagram 5

of a gas turbine power plant 7
with reheat and intercooling
6
4 8
2
3 1
Example 6.13 A gas turbine power plant operates with air of 14.7 psi and 50
deg_F,
√ entering the low-pressure compressor. This compressor pressure ratio
is 10 and there is intercooling at its outlet back to the initial temperature.
Subsequently, there is a high pressure compressor that works with the same
pressure ratio. The high-pressure turbine inlet temperature is 2550 ◦ F and it
works at the same pressure ratio as the compressors. At its outlet there is
another combustor that increases the air temperature back to 2550 ◦ F. Finally,
the expansion through the low-pressure turbine is back to the inlet pressure
to the first compressor. Calculate the net specific work, back work ratio, heat
transfer, and thermal efficiency.
Solution.
√
rp = 10
State 1 p1 = 14.7 psi Tf 1 = 50◦ F
h1 (Tf 1 ) = 12.037 Btu
lbm pr1 (Tf 1 ) = 1.4370
State 2 p 2 = p 1 rp pr2 = pr1 rp = 4.5442
f2 (pr2 ) = 0.8518 h2 (f2 ) = 59.560 Btu
lbm wc1 = h1 − h2 = −47.523 Btu
lbm
State 3 p3 = p2 Tf 3 = Tf 1
h3 = h1 pr3 = pr1
State 4 p 4 = p 3 rp pr4 = pr3 rp = pr2
h4 = h2 wc2 = h3 − h4 = wc1 wc = 2wc1 = −95.05 Btu
lbm
State 5 p5 = p4 Tf 5 = 2550◦ F
h5 (Tf 5 ) = 688.17 Btu
lbm pr5 (Tf 5 ) = 1268.1 qh1 = h5 − h4 = 628.61 Btu
lbm
p5 pr5
State 6 p6 = rp pr6 = rp = 401.01
f6(pr6 ) = 0.6343 h6 (f6 ) = 479.98 Btu
lbm wt1 = h5 − h6 = 208.19 Btu
lbm
h7 = h5 pr7 = pr5 qh2 := h7 − h6 = 208.19 Btu
lbm
State 8 p8 = p1 pr8 = prr7
p
= p r6
h8 = h6 wt2 = h7 − h8 = wt1 wt = 2wt1 = 416.38 Btu

lbm
Cycle Quantities
w = wt + wc = 321.33 Btu
lbm qh = qh1 + qh2 = 836.80 Btu
lbm
|wc | w
bwr = wt = 22.8% ηT = qh = 38.4%
Comparing these results with those of Example 6.12, we see that reheat has increased
the turbine work while intercooling has reduced the input compressor work. As a
result, the back work ratio has decreased. However, due to a larger fractional increase
in qh than in w, the thermal efficiency has decreased. The heat transfer increase, and
resulting thermal efficiency decrease, can be reversed by including a regenerator in
the cycle.
Regeneration
This concept makes use of the high-temperature, high-energy exhaust gas exiting the
turbine to preheat the air exiting the compressor, thereby reducing the amount of fuel
needed to raise its temperature to the desired turbine inlet value. It is implemented by
using a counterflow heat exchanger, the regenerator, which uses hot turbine exhaust
gases to heat cold air from the compressor exhaust.
The regenerator analysis is based on knowledge of a single parameter, the regener-
ator efficiency, which in terms of the numbered states in the cycle shown in Figs. 6.16
and 6.17 is
T10 − T4
ηreg =
T8 − T4
This varies between 0, when T10 = T4 , and 1, when T10 = T8 . This equation specifies
T10 , in this cycle, when ηreg is specified. Subsequently, the enthalpy exhausted to the
heat rejector is given by the energy equation for the regenerator (see this in Fig. 6.16).
Adding an 80% efficient regenerator to our previous example gives the following in
which states 1 through 8 are the same as the previous example while state 10 is
calculated from the regenerator efficiency relation.
Fig. 6.16 The flow diagram of a gas turbine power plant with reheat, intercooling, and regeneration
Fig. 6.17 The T -s diagram

of a gas turbine power plant
with reheat, intercooling, and
regeneration
Example 6.14 A gas turbine power plant operates with air of 14.7 psi and
50 deg_F,√ entering the low-pressure compressor. This compressor pressure
ratio is 10 and there is intercooling at its outlet back to the initial temper-
ature. Subsequently, there is a high-pressure compressor that works with the
same pressure ratio. The high-pressure turbine inlet temperature is 2550 ◦ F
and it has the same pressure ratio as the compressors. At its outlet, there is
another combustor that increases the air temperature back to 2500 ◦ F. Finally,
the expansion through the low-pressure turbine is back to the inlet pressure
of the first compressor. A regenerator of 80% efficiency is connected to the
exhaust of this low-pressure turbine and is used to preheat the air entering the
first combustor. Calculate the net specific work, back work ratio, heat transfer,
and thermal efficiency.
Solution.
√
rp = 10
State 1 p1 = 14.7 psi Tf 1 = 50◦ F
h1 (Tf 1 ) = 12.037 Btu
lbm pr1 (Tf 1 ) = 1.4370
State 2 p 2 = p 1 rp pr2 = pr1 rp = 4.5442
f2 (pr2 ) = 0.8518 h2 (f2 ) = 59.560 Btu
lbm Tf 2 (f2 ) = 246.3◦ F
wc1 = h1 − h2 = −47.52 Btu
lbm
h3 = h1 pr3 = pr1
State 4 p4 = p3 rp pr4 = pr3 rp = pr2
Tf 4 = Tf 2 h4 = h2 wc2 = h3 − h4 = wc1
State 5 p5 = p4 Tf 5 = 2550◦ F
h5 (Tf 5 ) = 688.17 Btu
lbm pr5 (Tf 5 ) = 1268.1
p5 pr5
State 6 p6 = rp = p2 pr6 = rp = 401.01
f6 (pr6 ) = 0.6343 h6 (f6 ) = 480.00 Btu
lbm Tf 6 (f6 ) = 1831.7◦ F
wt1 = h5 − h6 = 208.19 Btu
lbm
h7 = h5 pr7 = pr5 qh2 = h7 − h6 = wt1
pr7
State 8 p8 = p1 pr8 = rp = pr6
Tf 8 = Tf 6 h8 = h6 wt2 = h7 − h8 = wt1
State 10 p10 = p4 Tf 10 = Tf 4 + ηreg (Tf 8 − Tf 4 )
Tf 10 == 1514.6◦ F h10 (Tf 10 ) = 390.63 Btu
lbm qh1 = h5 − h10 = 297.48 Btu
lbm
Cycle Quantities
wc = wc1 + wc2 = −95.05 Btu Btu Btu
lbm wt = wt1 + wt2 = 416.38 lbm w = wt + wc = 321.33 lbm
|wc |
bwr = wt = 22.8% qh = qh1 + qh2 = 505.67 Btu
lbm ηT =
w
qh = 63.5%
Fig. 6.18 The T -s diagram of a gas turbine power plant showing the compressor and turbine losses
We did not need to calculate anything about state 9 although we could have,
using the energy equation for the regenerator. The regeneration increased the inlet
temperature to the high-pressure turbine and consequently reduced the heat transfer
required in that burner and thereby increased the thermal efficiency to 63.5%.
Compressor and Turbine Losses

As we have noted before, in the Brayton cycle compressor work is a significant
fraction of the turbine work. This effect is measured by the back work ratio; defined
in Eq. (6.15). Since losses in the compressor and the turbine both act to decrease the
back work ratio, they should be incorporated into our analysis.
The relevant calculations have already been described in Sect. 6.2.3, Turbine and
Pump Losses. Here we adapt Example 6.12 to this calculation. We have shown the
relevant T -s diagram in Fig. 6.18.
Example 6.15 A gas turbine power plant operates with air of 14.7 psi, 50 deg_F
entering the compressor. The compressor pressure ratio is 10 and the turbine
inlet temperature is 2550 ◦ F. The efficiencies of the turbine and compressor are
87% and 85% respectively. Calculate the net specific work, back work ratio,
heat transfer, and thermal efficiency. In addition, calculate the mass rate of flow
and the power developed, knowing that the inlet diameter of the compressor
is 2 ft.
Solution.
rp = 10 ηt = 87% ηc = 85%
State 1 p1 = 14.7 psi Tf 1 = 50◦ F
h1 (Tf 1 ) = 12.037 Btu
lbm pr1 (Tf 1 ) = 1.4370
State 2s p2s = p1 rp pr2s = pr1 rp = 14.370
f2s (pr2s ) = 0.5550 h2s (f2s ) = 125.31 Btu
lbm
wc∗
wc∗ = h1 − h2s = −113.28 Btu
lbm wc = ηc = −133.27 Btu
lbm
State 2 p2 = p2s h2 = h1 − wc = 145.30 Btu
lbm
State 3 p3 = p2 Tf 3 = 2550◦ F
h3 (Tf 3 ) = 688.17 Btu
lbm pr3 (Tf 3 ) = 1268.1 qh = h3 − h2 = 542.87 Btu
lbm
p3 pr3
State 4s p4s = rp = p1 pr4s = rp = 126.81
f4s (pr4s ) = 0.3291 h4s (f4s ) = 322.29 Btu
lbm wt∗ = h3 − h4s = 365.88 Btu
lbm
wt = wt∗ ηt = 318.32 Btu
lbm
Cycle Quantities
w = wt + wc = 185 Btu
lbm bwr= |w c|
wt = 41.9% ηT = w
qh = 34.1%
There is no need to calculate state 4; h 4 is not needed since wt is already calculated.

These results show that, as a result of the compressor and turbine losses, the
back work ratio is increased and the power developed is decreased. Even though the
heat transfer to the cycle is reduced, the fractional amount is less than the fractional
reduction in the net work done and so the thermal efficiency is also reduced by the
losses.
The Jet Engine
The jet engine is a modification of the gas turbine in which the energy contained in
the fuel is converted into the kinetic energy of the stream of burned exhaust gases.
This corresponds to a change in momentum of the gas flowing through the engine
which, by Newton’s law of motion, corresponds to a force acting on the engine in the
direction of the flowing gas. A flow diagram of the device is shown in Fig. 6.19 and
we can see there that there is a diffuser at the inlet to the compressor and a nozzle at
the exit of the turbine (see Sect. 4.4.2, Duct Flow for a discussion of diffusers and
nozzles).
The purpose of these devices is to convert kinetic energy to and from internal
energy. The turbine is designed to produce the exact amount of work required to
drive the compressor. In terms of the state numbering shown on the T -s diagram of
the figure, the engine thrust is
F = Ṁ(v6 − v1 ) (6.19)
Kinetic energy is only significant at the inlet of the diffuser and the exit of the nozzle.
Everywhere else in the engine, we neglect it.
Fig. 6.19 The flow diagram and corresponding T -s plot of an aircraft gas turbine
The following example illustrates the procedure for solving jet engine problems.
Note that we use the condition wt + wc = 0 to specify the enthalpy at the turbine
outlet. This is framed in the example to highlight it (Fig. 6.20).
Example 6.16 Air enters a turbojet engine at 13.4 psi and 50 ◦ F (an altitude of
2500 ft) with an inlet speed of 300 mph. The pressure ratio across the com-
pressor is 8, the turbine inlet temperature is 2000 ◦ F, and the mass rate of flow
through the engine is 90 lbm/s. Calculate the exit speed of the exhaust gas, and
the engine thrust.
Solution.
Ṁ = 90 lbm
s rp = 8
State 1 p1 = 12.3 psi Tf 1 = 50◦ F
h1 (Tf 1 ) = 12.037 Btu
lbm pr1 (Tf 1 ) = 1.4370 v1 = 300 mph
v12
State 2 s2 = s1 h2 = h1 + 2 = 15.903 Btu
lbm
pr2
f2 (h2 ) = 0.3600 pr2 (f2 ) = 1.6062 p2 = pr1 p1 = 14.98 psi
State 3 p3 = p2 rp = 119.8 psi pr3 = pr2 rp = 12.850
f3 (pr3 ) = 0.6249 h3 (f3 ) = 117.97 Btu
lbm wc = h2 − h3 = −102.07 Btu
lbm
State 4 p4 = p3 Tf 4 = 2000◦ F
h4 (Tf 4 ) = 528.08 Btu
lbm pr4 (Tf 4 ) = 538.50 qh = h4 − h3 = 410.11 Btu
lbm
State 5 s5 = s4 h5 = h4 + wc = 426.01 Btu
lbm
pr5
f5 (h5 ) = 0.1822 pr5 (f5 ) = 280.09 p5 = pr4 p4 = 62.32 psi
State 6 p6 = p1 pr6 = pp65 pr5 = 60.221
f6 (pr6 ) = 0.7194 h6 (f6 ) = 242.08 Btu
lbm v6 = 2(h5 − h6 ) = 3035 fts
Cycle Quantities
F = Ṁ (v6 − v1 ) = 7259 lbf
Fig. 6.20 The thrust and propulsive power by a jet engine as a function of flight speed, v1 . The
equilibrium flight speed, v1e , corresponds to the value for which the thrust and drag, Fd , forces are
equal
Note that the kinetic energy terms in the energy equation for each device are
neglected except for those of States 1 and 6.
If the engine is located in a plane flying at the speed v1 , the quantity
P = Fv1 (6.20)
is the propulsive power. Note that regarded as functions of the aircraft speed, the
thrust and power are represented by the functions sketched in Fig. 6.3.
6.4 Refrigeration
As we pointed out in Sect. 5.5.4 a refrigerator is a heat engine operated in reverse.

We use refrigeration cycles today to help preserve food and for our personal comfort;
not only for cooling, but, as heat pumps, also for heating as well. You will find a
discussion of refrigerators on Sect. 5.5.4, Refrigerators.
In this section, we will explore some of the schemes that have been devised to
implement refrigeration. The unit used to describe refrigerating capacity is the Ton
Btu
1 Ton = 12, 000 (6.21)
hr
As you can see from the definition it is a power. Like the horsepower that was devised
by James Watt for marketing reasons, this unit was introduced in connection with the
marketing of ice making machines; it is the rate of heat transfer required to freeze
(approximately) one ton (2000 lbm) of water at 32 ◦ F in 24 h. We denote the ton of
mass by a lower case t and the Ton of refrigeration by an upper case T to distinguish
them.
6.4 Refrigeration 267
As we discussed in Sect. 5.5.4, there are three quantities of importance in refriger-

ation; these are the rate of heat transfer, Q̇ L , (this is the refrigerating capacity) to the
refrigerator from the cold space (the volume to be refrigerated at temperature TL ),
the power, Ẇ M supplied to the refrigerator by an external motor, and the rate of heat
transfer, Q̇ H , from the refrigerator to a hot reservoir (the large volume surrounding
the refrigerator at temperature TH > TL ). As we learned in Sect. 5.5.4, in the steady
state, these quantities are related by the first and second laws and given by Eq. (5.58)

TH
Ẇ M ≥ Q̇ L −1
TL
Given a refrigerating capacity and the two working absolute temperatures, this
equation bounds the power required from below. The coefficient of performance,
Eq. (5.59),
Q̇ L
COP R =
Ẇ M
is used to compare the relative efficiency of two refrigerators that have different
capacities and use different amounts of power. However, it is not called an efficiency
because its value is larger than one.
∗
The minimum power required, Ẇ M , corresponds to the equality in Eq. (5.58); the
device that does this is called a Carnot refrigerator. The thermodynamic efficiency,
also called the second law efficiency is given by
∗
Ẇ M
η=
Ẇ M
For a refrigerator in which the working substance circulates in a closed loop at a rate
given by Ṁ, specific quantities
Ẇ M Q̇ H Q̇ L
wM = qH = qL =
Ṁ Ṁ Ṁ
are used.
6.5 Vapor Refrigeration Cycle
A vapor compression refrigerator operates by using the heat transfer from the cold
space to evaporate the working fluid, then compresses the vapor to a higher pressure,
condenses the high-pressure vapor by heat transfer to the surroundings, and finally
returns the liquid to its original pressure by throttling it.
The working fluid circulates around the loop composed of these four devices at
a rate Ṁ which is the same in each steady flow device. A suitable working fluid
has a liquid–vapor phase change in the temperature range of operation, is nontoxic,

environmentally friendly, and readily available. We will use in our examples a spe-
cific refrigerant, R-12 although modern devices use more environmentally friendly
refrigerants.
6.5.1 The Ideal Vapor Compression Cycle
Both a flow diagram and T -s plot for an ideal system are shown in Fig. 6.21. The
ideal system is characterized by having the inlet to an isentropic compressor as a
saturated vapor and the outlet of the condenser as a saturated liquid. This means
that the cycle is completely specified by the evaporator and condenser pressures (or
temperatures). Also shown on the T -s plot are the two reservoir temperatures. Note
that TH must be less than the condenser temperature while TL must be greater than
the evaporator temperature. The following example illustrates the calculations we
can make. In the example, f 2 is an interpolation variable on entropy and h 2 is the
corresponding interpolated value for enthalpy from values in the table for R-12.
Example 6.17 Calculate the COP, the mass rate of flow, and power required
for 1.5 Ton of refrigeration in an ideal refrigeration cycle operating between
an evaporator pressure of 20 psi and a condenser pressure of 100 psi and using
R-12 as the working fluid. If the reservoir temperatures are 70 ◦ F and 10 ◦ F,
calculate COP∗ and the thermodynamic efficiency (Sect. 5.5.5, Two Reservoir
Systems) of the refrigerator.
Solution.
State 1 p1 = 20 psi x1 = 1
h1 (p1 , x1 ) = 76.78 Btu
lbm
Btu
s1 (p1 , x1 ) = 0.17055 lbm R
State 2 s2 = s1 p2 = 100 psi
0.17055−0.16517
f2 (s2 , p2 ) = 0.17110−0.16517 = 0.9073 h2 (f2 , p2 ) = [86.029 + f2 (89.287 − 86.029)] Btu
lbm
h2 (f2 , p2 ) = 88.985 Btu
lbm wM = h2 − h1 = 12.207 Btu
lbm
State 3 p3 = p2 x3 = 0
h3 (p3 , x3 ) = 26.818 Btu
lbm
State 4 h4 = h3 p4 = p1
ql = h1 − h4 = 49.960
Cycle Quantities
COP= wqMl = 4.10 Ṁ = 1.5 Ton lbm
ql = 6.00 min
ẆM = wM Ṁ = 1.73 hp COP∗ = TL
= 7.83 η = COP∗ = 52.4%
TH −TL COP
6.5 Vapor Refrigeration Cycle 269
Fig. 6.21 A typical vapor compression refrigerator has four component devices; its T -s plot looks
similar to that of a Rankine cycle, but it is traversed in reverse (counterclockwise)
Fig. 6.22 A heat exchanger that uses subcooling the liquid to superheat the vapor in a vapor
compression refrigeration cycle
Notice how the mass rate of flow is specified by the rated capacity, Q̇ L = 1.5 Tons,
of the unit. Note also that T f 3 ≥ T f H and T f 1 ≤ T f L as required by the irreversibility
of heat transfer (the second law of thermodynamics).
6.5.2 Subcooling and Superheating
It is desirable in practice for state 3 in Fig. 6.21 to be in the liquid region, rather than
on the saturation curve, in order to prevent the possibility of getting vapor into the
throttle. Likewise state 1 should be in the vapor region, rather than on the saturation
curve in order to prevent the possibility of liquid getting into the compressor. Both
subcooling and superheating can be accomplished by adding, one device to the cycle,
a heat exchanger as shown in Fig. 6.22, and analyzed in the following example.
Example 6.18 Calculate the COP, the mass rate of flow, and the power required
for 1.5 Ton of refrigeration in a refrigeration cycle has a compressor inlet of
20 psi, 0 ◦ F, a condenser inlet of 100 psi and uses R-12 as the working fluid.
The evaporator and condenser are constant pressure heat exchangers, whose
outlet states are saturated vapor and liquid, respectively. The compressor is
isentropic and there is a heat exchanger as shown in Fig. 6.22 to subcool the
liquid. If the reservoir temperatures are 70 ◦ F and 10 ◦ F, calculate COP∗ and
the thermodynamic efficiency of the refrigerator.
Solution.
State 6 p6 = 20 psi x6 = 1
h6 (p6 , x6 ) = 76.778 Btu
lbm
State 1 p1 = p6 Tf 1 = 0◦ F
h1 (p1 , Tf 1 ) = 77.929 Btu
lbm s1 (p1 , Tf 1 ) = Btu
0.017307 lbm R
State 2 s2 = s1 p2 = 100 psi
f2 (s2 , p2 ) = 0.3362 h2 (f2 , p2 ) = 90.411 Btu
lbm wM = h2 − h1 = 12.6 Btu
lbm
State 3 p3 = p2 x3 = 0
h3 (p3 , x3 ) = 26.818 Btu
lbm
State 4 p4 = p3 h4 = h3 + h6 − h1 = 25.667 Btu
lbm
State 5 h5 = h4 p5 = p1
ql = h6 − h5 = 51.111 Btu
lbm
Cycle Quantities
ql = 5.870 min
ẆM = wM Ṁ = 1.73 hp COP∗ = TL
= 7.83 η = COP∗ = 52.2%
TH −TL COP
There is little difference between these results and those of the cycle without
subcooling and superheating in Example 6.17. Here the power required is a bit larger
while the mass flow is a little smaller, and while the efficiency is a bit smaller, there
is no chance of vapor entering the throttle or liquid entering the compressor.
6.5.3 Compressor Loss
The major loss in a vapor compression refrigeration system is the compressor loss
just as it was in power plants, and we treat this aspect in the same way here that we
did there.
6.5 Vapor Refrigeration Cycle 271
Example 6.19 Calculate the COP, the mass rate of flow, and the power required
for 1.5 Ton of refrigeration in a refrigeration cycle operating between an evap-
orator pressure of 20 psi and a condenser pressure of 100 psi and using R-12 as
the working fluid. The evaporator and condenser are constant pressure devices
whose outlet states are saturated vapor and liquid, respectively, but the com-
pressor is only 85% efficient. If the reservoir temperatures are 70 ◦ F and 10 ◦ F,
calculate COP∗ and the thermodynamic efficiency of the refrigerator.
Solution.
ηc = 85%
State 1 p1 = 20 psi x1 = 1
h1 (p1 , x1 ) = 76.778 Btu
lbm s1 (p1 , x1 ) = Btu
0.17055 lbm R
State 2s s2s = s1 p2s = 100 psi
f2s (s2s , p2s ) = 0.9073 h2s (f2s , p2s ) = 88.985 Btu
lbm
∗
wM
∗ = h − h = 12.207 Btu
wM wM = = 14.361 Btu
2s 1 lbm ηc lbm
State 2 p2 = p2s h2 = h1 + wM = 91.139 Btu
lbm
State 3 p3 = p2 x3 = 0
h3 (p3 , x3 ) = 26.82 Btu
lbm
State 4 h4 = h3 p4 = p1
ql = h1 − h4 = 49.96
Cycle Quantities
ql = 6.00 min
ẆM = wM Ṁ = 2.03 hp COP∗ = TL
TH −TL = 7.83 η = COP∗ = 44.4%
COP
As a result of the compressor loss, the power, COP, and η have increased,
decreased, and decreased, respectively, from their values in Example 6.17.
6.6 Exercises
Section 6.2.3: The Ideal Rankine Cycle
6.1 An ideal Rankine cycle has a boiler temperature of 150 ◦ C and a condenser
temperature of 60 ◦ C. The net power output of the cycle is 50 kW. What are the
boiler and condenser pressures? Calculate the steam mass rate of flow, the rate of
heat transfer to the water in the boiler, the thermal efficiency, and the steam quality
at the turbine outlet. Calculate the efficiency of a Carnot cycle with the same boiler
and condenser temperatures and then calculate the thermodynamic efficiency of the
Rankine cycle. (475.8 kPa, 19.94 kPa, 0.1013 kg/s, 252.7 kW, 19.8%, 84.9%, 21.3%,
93.0%)
6.2 Repeat the previous exercise using R-12 as the cycle working fluid, but with a
boiler temperature of 100 ◦ C.
Rankine cycle. (3.973 MPa, 19.94 kPa, 0.0625 kg/s, 159.2 kW, 31.4%, 74.0%, 36.3%,
86.5%)
Rankine cycle.
Section 6.2.3: Superheat
6.5 A Rankine cycle operates with a turbine inlet state of 600 kPa and 300 ◦ C and
a condenser temperature of 60 ◦ C. The condenser outlet is saturated liquid at the
turbine outlet pressure and the pump is isentropic with outlet pressure the same as
the turbine inlet. Calculate for a 50 kW plant, the steam mass rate of flow, the rate of
at the turbine outlet. Assuming that the cooling water temperature rise is 15 ◦ C,
calculate the cooling water mass rate of flow through the condenser (consider the
cooling water a constant property fluid). (0.0793 kg/s, 222.8 kW, 22.4%, 92.4%, 2.75
kg/s)
6.6 A Rankine cycle operates with a turbine inlet state of 1.4 MPa and 500 ◦ C. Take
the condenser outlet as saturated liquid at 20 ◦ C and calculate for a 50 kW plant,
the steam mass rate of flow, the rate of heat transfer to the water in the boiler, the
thermal efficiency, and the steam temperature at the turbine outlet. Assuming that
the cooling water temperature rise is 15 deg_C, calculate the cooling water mass rate
of flow through the condenser (consider the cooling water a constant property fluid).
6.7 A Rankine cycle operates with a turbine inlet temperature of 500 ◦ C and turbine
outlet quality of 95%. If the condenser temperature is 20 ◦ C and its outlet is saturated
liquid, calculate the boiler pressure, specific turbine work, pump work, boiler heat
transfer, and thermal efficiency. Start your calculations with the turbine outlet state.
(359 kPa, 1061 kJ/kg, 0.36 kJ/kg, 3392 kJ/kg, 31.3%)
6.6 Exercises 273
Section 6.2.3: Reheat
6.8 A Rankine cycle with one stage of reheat operates with its high-pressure turbine
inlet at 500 ◦ C, 10 MPa and its low- pressure turbine inlet at 500 ◦ C, 400 kPa. If
the condenser temperature is 20 ◦ C and its outlet is saturated liquid calculate the,
quality at the high-pressure and low-pressure turbine outlets, turbine work, pump
work, total boiler heat transfer, and thermal efficiency. (94.2%, 94.2%, 1840.1 kJ/kg,
10.02 kJ/kg, 4142.2 kJ/kg, 44.2%)
6.9 A Rankine cycle with one stage of reheat operates with its high-pressure turbine
inlet at 800 ◦ F and outlet states at 95% quality. The reheat pressure is 20 psi, the
condenser temperature is 60 ◦ F and condenser outlet is saturated liquid. Calculate the
boiler pressure, reheat temperature, turbine work, pump work, boiler heat transfer,
and thermal efficiency. Start your calculations at the low-pressure turbine outlet,
state 4.
Section 6.2.3 Regeneration
6.10 Modify the cycle of Exercise 6.5 to include regeneration with an open feed-
water heater as shown in Fig. 6.8. The extraction pressure is 100 kPa. Calculate
the extraction fraction, turbine work, pump work, boiler heat transfer, and thermal
efficiency. (6.84%, 614.0 kJ/kg, 0.60 kJ/kg, 2643.6 kJ/kg, 23.2%)
6.11 Modify the cycle of Exercise 6.5 to include regeneration with a closed feedwater
heater as shown in Fig. 6.9. The extraction pressure is 100 kPa, and the temperature
at the boiler input is 90 ◦ C. Calculate the extraction fraction, turbine work, pump
work, boiler heat transfer, and thermal efficiency.
Section 6.2.3: Turbine and Pump Losses
6.12 Modify the results of Exercise 6.5 to include a pump efficiency of 85% and a
turbine efficiency of 87%. (0.0912 kg/s, 256.2 kW, 19.5%, 95.9%, 3.28 kg/s)
6.13 Modify the results of Exercise 6.6 to include a pump efficiency of 85% and a
turbine efficiency of 87%.
Section 6.3.1
6.14 An Otto cycle operates from inlet (State 1) conditions of 100 kPa and 25 ◦ C, a
compression ratio of 8, and a qh , the specific heat transfer released in the combustion
process, of 1250 kJ/kg air. Calculate the pressures and absolute temperatures at the
other state points of the cycle as well as the net specific work, thermal efficiency,
and mep. Compare the thermal efficiency with the value from Eq. (6.13). What is
the cause of the difference? (1.50 MPa, 560 K, 5.20 MPa, 1940 K, 232 kPa, 692 K,
696 kJ/kg, 55.7%, 929 kPa, 56.5%)
6.15 An Otto cycle operates from inlet (State 1) conditions of 14.5 psi and 50 ◦ F, a
compression ratio of 8, and a qh , the specific heat transfer released in the combustion
process, of 550 Btu/lbm air. Calculate the net specific work, thermal efficiency, and
mep. Compare the thermal efficiency with the value from Eq. (6.13). What is the
cause of the difference?
6.16 In a 4-cylinder engine, the cylinder bore is 3.7 in and stroke is 3.5 in. The
clearance volume is 12% of the volume at bottom dead center and the crankshaft is
turning at 3000 rpm. The state 1 conditions are 14.5 psi and 50 ◦ F and the highest
cylinder temperature is 4550 ◦ F, calculate the net specific work, net work, power,
specific heat transfer from the fuel, thermal efficiency, and mep. (427.9 Btu/lbm,
2532 ft lbf, 115 hp, 858.6 Btu/lbm, 49.8%, 202 psi)
Section 6.3.2
6.17 The conditions at the beginning of compression in a Diesel cycle are 100 kPa
and 25 ◦ C. The compression ratio is 20 and the specific heat transfer from the com-
bustion process is 900 kJ/kg. Determine the maximum absolute temperature, the
maximum pressure, the cutoff ratio, the net specific work, the thermal efficiency, and
the mep.
6.18 The displacement volume of a Diesel engine is 3 liters and the cutoff ratio
is 2.5. The state of the air at the beginning of compression is 95 kPa, 25 ◦ C, in a
volume of 3.15 liters. Determine the compression ratio, the net work, and the power
developed if the cycle is executed 1500 times per minute. What is the rotational speed
of the crankshaft? (21, 961.7 kJ/kg, 84.1 kW, 3000 rpm)
6.19 The conditions at the beginning of compression in a Diesel cycle are 14.7
psi and 60 ◦ F. The compression ratio is 20 and the specific heat transfer from the
combustion process is 1000 Btu/lbm. Determine the maximum absolute temperature,
the maximum pressure, the cutoff ratio, the net specific work, the thermal efficiency,
and the mep.
Section 6.3.3
6.20 Air enters the compressor of an air standard Brayton cycle at 100 kPa, 25 ◦ C
and a volumetric flow rate of 5 m3 /min. The compressor pressure ratio is 10 and the
turbine inlet temperature is 1325 ◦ C. Calculate the net work, net power developed,
back work ratio, and thermal efficiency of the cycle. (535 kJ/kg, 52.1 kW, 34.1%,
45.0%)
6.21 Derive Eq. (6.16) using the equations of state for a perfect gas. Show that the
back work ratio, Eq. (6.15), can be written as bwr= T1 /T4 where T1 is the compressor
inlet absolute temperature and T4 is the turbine outlet absolute temperature.
6.22 Air enters the compressor of an air standard Brayton cycle at 14.7 psi, 75 ◦ F
and a volumetric flow rate of 150 ft3 /min. The compressor pressure ratio is 10 and the
turbine inlet temperature is 2300 ◦ F. Calculate the net work, net power developed,
back work ratio, and thermal efficiency of the cycle. (216 Btu/lbm, 56.6 hp, 35.5%,
45.1%)
6.6 Exercises 275
Section 6.3.3: Reheat and Intercooling
6.23 A two-stage air compressor operates at steady state, compressing 5 m3 /min of

air isentropically from 100 kPa, 25 ◦ C to 1 MPa. An intercooler between the two
stages cools the√air back to the initial Celsius temperature at a constant intermediate
pressure of 100 10 kPa. Calculate the power required to run the compressors and
compare the result to the power required to compress the air isentropically, in one
stage, from the given initial state to the final pressure. (22.6 kW, 27.0 kW)
6.24 Using the perfect gas equations of state show that the intermediate pressure for
√
minimum compressor work when T3 = T1 is given by p2 = p4 p1 and in that case
the minimum work is wmin = 2c p T1 [1 − ( p4 / p1 )(k−1)/2k ].
6.25 The two compressors of Exercise 6.23 are connected by a constant pressure
combustor to a two-stage turbine setup (see Fig. 6.14). The inlet temperature to the
high-pressure
√ turbine is 1400 ◦ C and it expands the burned gases isentropically to
100 10 kPa. These gases are then piped to another constant pressure combustor
where they are heated again to 1400 K. They are then expanded isentropically in the
low-pressure turbine back to the original inlet pressure. Calculate the net work, net
power developed, back work ratio, and thermal efficiency of the cycle. (737 kJ/kg,
71.8 kW, 23.9%, 38.2%)
6.26 A two-stage air compressor operates at steady state, compressing 150 ft3 /min of
air isentropically from 14.7 psi, 50 ◦ F with a pressure ratio of 5 for each compressor.
An intercooler between the two stages cools the air back to the initial Fahrenheit
temperature at constant intermediate pressure. Calculate the power required to run
the compressors and compare the result to the power required to compress the air
isentropically, in one stage, from the given initial state to the final pressure.
6.27 The two compressors of the previous exercise are connected by a constant
pressure combustor to a two-stage turbine setup with the same intermediate pressure.
The inlet temperature to the high pressure turbine is 3000 ◦ F and it expands the
burned gases isentropically to the intermediate pressure. These gases are then piped
to another constant pressure combustor where they are heated again to 3000 R. They
are then expanded isentropically in the low pressure turbine back to the original inlet
pressure. Calculate the net work, net power developed, back work ratio, and thermal
efficiency of the cycle. (494.7 Btu/lbm, 136 hp, 22.3%, 46.8%)
Section 6.3.3: Regeneration
6.28 Add a regenerator with efficiency of 80% to the Brayton cycle of Exercise 6.25
and recalculate the net work, net power developed, back work ratio, and thermal
efficiency of the cycle. (737 kJ/kg, 71.8 kW, 23.9%, 62.8%)
6.29 Add a regenerator with efficiency of 70% to the Brayton cycle of Exercise 6.27
and recalculate the net work, net power developed, back work ratio, and thermal
efficiency of the cycle.
Section 6.3.3: Compressor and Turbine Losses
6.30 Let the compressor efficiency be 85% and the turbine efficiency be 87% in
Exercise 6.20 and redo all the calculations. (380 kJ/kg, 37.0 kW, 46.1%, 33.4%)
6.31 Let the compressor efficiency be 85% and the turbine efficiency be 87% in
Exercise 6.22 and redo all the calculations.
Section 6.3.3: The Jet Engine
6.32 Air at 22 kPa, −50 ◦ C, and 225 m/s enters a turbojet engine in flight (an
altitude of 10,000 m). The pressure ratio across the compressor is 12. The turbine
inlet temperature is 1523.15 K and the nozzle exit pressure is the same as at the inlet.
The diffuser, compressor, turbine, and nozzle are isentropic. With a mass rate of flow
through the engine of 35 kg/s calculate the nozzle exit velocity, the engine thrust,
and the propulsive power. (1136 m/s, 31.9 kN, 7171 kW)
6.33 Modify the previous problem so that the compressor is 85% efficient while the
turbine has an efficiency of 87%.
6.34 Air at 10.018 psi, 25 ◦ F, and 400 ft/s enters a turbojet engine in flight (an
altitude of 10000 ft). The pressure ratio across the compressor is 8. The turbine inlet
temperature is 1800 ◦ F and the nozzle exit pressure is the same as at the inlet. The
diffuser, compressor, turbine, and nozzle are isentropic. With a mass rate of flow
through the engine of 90 lbm/s calculate the nozzle exit velocity, the engine thrust,
and the propulsive power. (2865 ft/s, 6896 lbf, 5015 hp)
Section 6.5.1
6.35 A refrigerator uses R-12 as its working fluid and operates on an ideal vapor
compression cycle between 0.15 and 0.7 MPa. The mass rate of flow of refrigerant
is 1 kg/s. Determine the rate of heat transfer to the cold box, the power input, rate
of heat transfer to the surroundings, and the COP. (116.5 kW, 27.2 kW, 143.8 kW,
4.28)
6.36 An ice maker uses R-12 as its working fluid and operates on an ideal vapor
compression cycle between 20 and 100 psi. Determine the power input required to
produce 15 lbm/hr of ice given that a heat transfer of 169 Btu from liquid water is
needed to freeze 1 lbm of it and the COP.
Section 6.5.2
6.37 R-12 enters the isentropic compressor of a refrigerator as superheated vapor

at 0.15 MPa −10 ◦ C and a rate of 1 kg/s; it leaves at 0.7 MPa. It is cooled in the
condenser to 20 ◦ C with no pressure loss. Determine the rate of heat transfer to the
cold box, the power input, rate of heat transfer to the surroundings, and the COP.
(130.1 kW, 30.2 kW, 160.3 kW, 4.31)
6.6 Exercises 277
6.38 Consider the ice maker of exercise 6.36 and insert a closed heat exchanger that
subcools the condenser output to 60 ◦ F while superheating the evaporator output
(see Fig. 6.22). Determine the resulting cycle COP and power required. Begin your
calculation at State 3, the condenser output.
Section 6.5.3
6.39 Modify exercise 6.35 so that the compressor has an efficiency of 85% and
calculate the changed values. (116.5 kW, 32.0 kW, 148.6 kW, 3.64)
6.40 Modify exercise 6.36 so that the compressor has an efficiency of 80% and
calculate the changed values.
Appendix A
Thermodynamic Properties: English Units

Mechanical Engineering Series, https://doi.org/10.1007/978-3-030-25221-2
280 Appendix A: Thermodynamic Properties: English Units
A.1 Linear Elastic Liquids
Material v0 α0 × 106 β0 × 106 cp cv

in3 /lbm 1/R in2 /lbf Btu/lbm R Btu/lbm R
Alcohol, ethyl 35.7 610 7.69 0.58 0.48
Benzene 31.2 770 6.49 0.42 0.26
Glycerin 21.7 280 1.53 0.58 0.52
Mercury 2.04 100 0.279 0.03 0.026
Machine Oil 30.3 410 5.29 0.40 0.35
Water 27.8 115 3.14 1.00 0.994
The equations of state of a linear elastic liquid are
v = v O + v O α0 (T − TO ) − v O β0 ( p − p O ) u = c p (T − TO )
and
h = c p (T − TO ) + v O ( p − p O ) s = c p ln(T /TO )
where v O , α0 , β0 , c p , and cv are constants listed in the table above for a few liquids,
and TO , p O are constant reference values. These constants are not all independent;
they satisfy the relation
c p = cv + TO v O α02 /β0
Notice that for alcohol the value of α0 is 610 × 10−6 R−1 , and that other values
should be read in a corresponding way. Since u, h, and s are relative quantities,
only differences have physical meaning so their equations of state should be used as
follows:
u 2 − u 1 = c p (T2 − T1 ) h 2 − h 1 = c p (T2 − T1 ) + v O ( p2 − p1 ) s2 − s1 = c p ln(T2 /T1 )
According to these equations, an isentropic change of state is isothermal, and cv is

not needed. This occurs because of the relative incompressibility of liquids.
The mechanical equation of state is most often used in the incompressible form
v ∼ vO
or when specific volume differences are desired in the linear elastic form
v2 − v1 = v O α0 (T2 − T1 ) − v O β0 ( p2 − p1 )
If actual values of v are needed, the appropriate values of TO and p O are 80 ◦ F and
14.7 psi.
Source: This data was obtained from several sources. The data is only representative, values vary
with temperature as well as other variables.
Appendix A: Thermodynamic Properties: English Units 281
A.2 Air as an Ideal Gas
The ideal gas law is pv = RT . Values for h, u, and s (0) , are tabulated from
T T
h= c p (TC ) dTC u = h − RT s (0) = c p (TC )/TC dTC φ̂
T0 T0
using the following expression for c p :

1.5 −1.5
T T
c p = 3.753 · 10−5 + 0.31981 − 3.614 +
180 180
−2 −3
T T Btu
7.306 − 5.249
180 180 lbm R
for T ≥ 459.67 R and c p = 0.2395 Btu/lbm R for T < 459.67 R. The reference
temperature, T0 , is 459.67 R, and the gas constant, R is 0.068549 Btu/lbm R (53.343
ft lbf/lbm R). The entropy difference between two equilibrium states is found from
the formula
s2 − s1 = s2(0) − s1(0) + R ln( p1 / p2 )
For isentropic process use the auxiliary functions
(0)
(T )−s (0) (TO )]/R T −[s (0) (T )−s (0) (TO∗ )]/R
pr (T ) = e[s or vr (T ) = e
T0∗
along with p2 / p1 = pr 2 / pr 1 or v2 /v1 = vr 2 /vr 1 . At temperatures less than

200 ◦ F, air can be considered a perfect gas with a high degree of accuracy.
The equation for c p was obtained from values for N2 (75% by mass), and O2 (25% by mass) given
in: Fundamentals of Classical Thermodynamics, G.J. Van Wylen and R.E. Sonntag, second edition,
John Wiley Co., New York, 1973, Table A.9, p. 683.
T Tf h u s (0) pr vr
R ◦F Btu/lbm Btu/lbm Btu/lbm R
359.67 −100 −23.950 −48.605 −0.058755 0.42439 419.44
384.67 −75 −17.962 −44.331 −0.042660 0.53669 354.73
409.67 −50 −11.975 −40.058 −0.027580 0.66875 303.18
434.67 −25 −5.9875 −35.784 −0.013393 0.82252 261.54
459.67 0 0 −31.510 0 1 227.50
484.67 25 6.0062 −27.218 0.012723 1.2039 199.24
509.67 50 12.037 −22.901 0.024855 1.4370 175.53
534.67 75 18.078 −18.572 0.036429 1.7014 155.53
559.67 100 24.127 −14.238 0.047483 1.9991 138.56
584.67 125 30.176 −9.9030 0.058057 2.3325 124.06
609.67 150 36.226 −5.5664 0.068190 2.7041 111.58
634.67 175 42.278 −1.2283 0.077919 3.1164 100.79
659.67 200 48.333 3.1130 0.087276 3.5722 91.394
684.67 225 54.393 7.4590 0.096292 4.0744 83.167
709.67 250 60.459 11.812 0.10499 4.6259 75.926
734.67 275 66.534 16.173 0.11341 5.2299 69.522
759.67 300 72.620 20.545 0.12155 5.8898 63.834
784.67 325 78.718 24.929 0.12945 6.6090 58.759
809.67 350 84.829 29.327 0.13712 7.3912 54.215
834.67 375 90.956 33.740 0.14457 8.2400 50.132
859.67 400 97.099 38.169 0.15182 9.1595 46.450
884.67 425 103.26 42.616 0.15889 10.154 43.121
909.67 450 109.44 47.081 0.16577 11.227 40.101
934.67 475 115.64 51.565 0.17249 12.383 37.355
959.67 500 121.85 56.069 0.17906 13.628 34.851
984.67 525 128.09 60.592 0.18548 14.965 32.564
1000.7 550 134.35 65.137 0.19175 16.400 30.470
1034.7 575 140.63 69.702 0.19789 17.937 28.548
1059.7 600 146.93 74.288 0.20391 18.583 26.781
1084.7 625 153.25 78.895 0.20981 21.342 25.153
1109.7 650 159.59 83.523 0.21559 23.219 23.652
1134.7 675 165.95 88.173 0.22126 25.222 22.265
1159.7 700 172.34 92.844 0.22682 27.355 20.981
1184.7 725 178.74 97.536 0.23229 29.625 19.791
1209.7 750 185.17 102.25 0.23766 32.039 18.686
1234.7 775 191.62 106.98 0.24293 34.602 17.659
1259.7 800 198.09 111.74 0.24812 37.322 16.704
1284.7 825 204.58 116.51 0.25322 40.205 15.814
1309.7 850 211.09 121.31 0.25824 43.259 14.984
1334.7 875 217.62 126.13 0.26318 46.491 14.208
1359.7 900 224.17 130.96 0.26804 49.908 13.483
1384.7 925 230.73 135.82 0.27283 53.518 12.805
1409.7 950 237.32 140.69 0.27754 57.329 12.169
1434.7 975 243.93 145.59 0.28219 61.349 11.574
1459.7 1000 250.56 150.50 0.28677 65.588 11.014
1509.7 1050 263.87 160.38 0.29573 74.751 9.9953
1559.7 1100 277.25 170.33 0.30445 84.890 9.0929
1609.7 1150 290.70 180.35 0.31294 96.081 8.2914
1659.7 1200 304.21 190.45 0.32121 108.40 7.5774
1709.7 1250 317.80 200.60 0.32927 121.93 6.9394
1759.7 1300 331.44 210.82 0.33714 136.76 6.3679
1809.7 1350 345.15 221.10 0.34482 152.98 5.8547
1859.7 1400 358.92 231.44 0.35233 170.67 5.3926
1909.7 1450 372.74 241.84 0.35966 189.95 4.9756
1959/7 1500 386.62 252.29 0.36684 210.91 4.5985
2009.7 1550 400.55 262.79 0.37386 233.65 4.2568
2059.7 1600 414.54 273.35 0.38073 258.30 3.9465
2109.7 1650 428.57 283.96 0.38746 284.95 3.6641
2159.7 1700 442.66 294.61 0.39406 313.74 3.4068
2209.7 1750 456.79 305.31 0.40053 344.78 3.1718
2259.7 1800 470.96 316.06 0.40687 378.21 2.9569
2309.7 1850 485.18 326.85 0.41309 414.15 2.7600
2359.7 1900 499.44 337.69 0.41920 452.75 2.5794
2409.7 1950 513.74 348.56 0.42520 494.15 2.4134
2459.7 2000 528.08 359.47 0.43109 538.50 2.2606
2509.7 2050 542.46 370.43 0.43688 585.94 2.1198
2559.7 2100 556.88 381.42 0.44257 636.64 1.9899
2609.7 2150 571.34 392.45 0.44816 690.75 1.8698
2659.7 2200 585.83 403.51 0.45366 748.46 1.7587
2709.7 2250 600.35 414.60 0.45907 809.93 1.6558
2759.7 2300 614.91 425.74 0.46439 875.33 1.5603
2809.7 2350 629.50 436.90 0.46963 944.86 1.4717
2859.7 2400 644.12 448.09 0.47479 1018.7 1.3893
2909.7 2450 658.77 459.32 0.47987 1097.1 1.3126
2959.7 2500 673.46 470.57 0.48488 1180.1 1.2412
3009.7 2550 688.17 481.86 0.48980 1268.1 1.1746
3059.7 2600 702.91 493.17 0.49466 1361.2 1.1124
3109.7 2650 717.68 504.51 0.49945 1459.7 1.0543
3159.7 2700 732.47 515.88 0.50417 1563.8 1.0000
3209.7 2750 747.29 527.27 0.50882 1673.6 0.94914
3259.7 2800 762.14 538.69 0.51341 1789.5 0.90150
3309.7 2850 777.01 550.14 0.51794 1911.7 0.85682
3359.7 2900 791.91 561.61 0.52241 2040.5 0.81489
3409.7 2950 806.83 573.10 0.52682 2176.0 0.77550
3459.7 3000 821.78 584.62 0.53117 2318.6 0.73848
3509.7 3050 836.74 596.16 0.53546 2468.5 0.70365
3559.7 3100 851.73 607.72 0.53971 2626.1 0.67086
3609.7 3150 866.75 619.31 0.54389 2791.5 0.63997
3659.7 3200 881.78 630.91 0.54803 2965.1 0.61084
3709.7 3250 896.83 642.54 0.55212 3147.2 0.58336
3759.7 3300 911.91 654.19 0.55615 3338.1 0.55741
3809.7 3350 927.00 665.85 0.56014 3538.1 0.53290
3859.7 3400 942.12 677.54 0.56408 3747.5 0.50972
3909.7 3450 957.25 689.25 0.56798 3966.7 0.48780
3959.7 3500 972.41 700.97 0.57183 4195.9 0.46705
4009.7 3550 987.58 712.72 0.57564 4435.6 0.44739
4059.7 3600 1002.8 724.48 0.57940 4686.0 0.42876
4109.7 3650 1018.0 736.26 0.58313 4947.6 0.41110
4159.7 3700 1033.2 748.06 0.58681 5220.6 0.39433
4209.7 3750 1048.4 759.88 0.59045 5505.6 0.37842
4259.7 3800 1063.7 771.71 0.59405 5802.8 0.36330
4309.7 3850 1079.0 783.56 0.59762 6112.6 0.34894
4359.7 3900 1094.3 795.43 0.60115 6435.5 0.33528
4409.7 3950 1109.6 807.31 0.60464 6771.8 0.32228
4459.7 4000 1124.9 819.21 0.60810 7122.0 0.30991
4509.7 4050 1140.3 831.13 0.61152 7486.4 0.29813
4559.7 4100 1155.6 843.06 0.61490 7865.6 0.28690
4609.7 4150 1171.0 855.00 0.61826 8259.9 0.27620
4659.7 4200 1186.4 866.96 0.62158 8669.8 0.26599
4709.7 4250 1201.8 878.94 0.62487 9095.8 0.25626
4759.7 4300 1217.2 890.93 0.62812 9538.3 0.24696
4809.7 4350 1232.6 902.93 0.63135 9997.8 0.23809
4859.7 4400 1248.1 914.95 0.63454 10475 0 22961
4909,7 4450 1263.5 926.99 0.63771 10970 0.22150
4959.7 4500 1279.0 939.03 0.64084 11403 0.21375
5009.7 4550 1294.5 951.09 0.64395 12016 0.20634
5059.7 4600 1310.0 963.17 0.64703 12568 0.19925
5109.7 4650 1325.5 975.25 0.65008 13140 0.19246
5159.7 4700 1341.0 987.35 0.65311 13733 0.18595
5209.7 4750 1356.6 999.47 0.65610 14346 0.17972
5259.7 4800 1372.1 1011.6 0.65907 14982 0.17375
5309.7 4850 1387.7 1023.7 0.66202 15640 0.16802
5359.7 4900 1403.3 1035.9 0.66494 16320 0.16253
5409.7 4950 1418.9 1048.0 0.66784 17025 0.15726
5459.7 5000 1434.5 1060.2 0.67071 17753 0.15220
5509.7 5050 1450.1 1072.4 0.67345 18506 0.14735
5559.7 5100 1465.7 1084.6 0.67638 19284 0.14269
5609.7 5150 1481.4 1096.8 0.67918 20088 0.13821
5659.7 5200 1497.0 1109.1 0.68196 20919 0.13390
5709.7 5250 1512.7 1121.3 0.68471 21777 0.12976
5759.7 5300 1528.4 1133.5 0.68745 22663 0.12578
A.3 Perfect Gases
The equations of state of a perfect gas are
R
pv = RT u = cv T = T
k−1
and
k
h = cpT = RT s = R ln[T 1/(k−1) v]
k−1
where R, c p , and cv are constants listed in the table, and k is the ratio of specific heats
c p /cv . The constants are not all independent, they satisfy the relation c p = cv + R.
The equations of state for u, h, and s should be used in difference form
u 2 − u 1 = cv (T2 − T1 ) h 2 − h 1 = c p (T2 − T1 ) s2 − s1 = R ln[(T2 /T1 )1/(k−1) (v2 /v1 )]
Two other expressions of the entropic equation of state, which are sometimes useful,
can be obtained by eliminating either v or T using the ideal gas law
s = R ln R + R ln[T k/(k−1) / p] s = −cv ln R + cv ln[ pv k ]
They are also used in difference form
s2 − s1 = R ln[(T2 /T1 )k/(k−1) ( p1 / p2 )] s2 − s1 = cv ln[( p2 / p1 )(v2 /v1 )k ]
In addition, it is often desirable to use the mechanical equation of state in dimen-

sionless form ( p2 / p1 )(v2 /v1 ) = T2 /T1 .
A gas can be considered perfect only over a limited temperature range. Here the
(constant) values of the specific heats correspond to the vicinity of 80 ◦ F.
A.4 Water
v ft3 /lbm, u Btu/lbm, h Btu/lbm, s Btu/lbm ◦ F
In the range shown
In the shaded region correspond- In the shaded region corresponding

ing to an incompressible liquid (this to an ideal gas (this appears in the
appears in the table to the right and table to the left and below the double
above the double line indicating the line indicating the phase change), the
phase change), the following rela- following relations hold, to within
tions hold, to within 1%, on a row 1%, 0.5% or 0.1% on a row (at con-
(at constant T ) stant T )
v = vf v2 = p1 v1 / p2
u = uf u2 = u1
h = h f + ( p − ps )v f h2 = h1
s = sf s2 = s1 + R ln( p1 / p2 )
The subscripted values are those of The value of R is 0.11021 Btu/lbm

◦
the saturated liquid; they also appear F, the lightest shading is 1%, and
to the right of the phase change line. the darkest is 0.1%.
Source: ASME Steam Tables, C.A. Meyer, R.B. McClintock, G.J. Silvestri, R.C. Spencer Jr., fifth
edition, The American Society of Mechanical Engineers, New York, 1983.
A.5 Linear Elastic Solids
Material v0 α × 106 E × 10−6 ν cp cv

in3 R−1 psi Btu Btu
lbm lbm R lbm R
Aluminum, 2024-T3 10.0 12.6 10.6 0.33 0.23 0.21
Copper 3.09 9.2 16.0 0.345 0.092 0.085
Glass, average 8.85 4.3 9.2 0.235 0.199 0.198
Iron 3.52 6.5 28.5 0.27 0.108 0.104
Lead, pure 2.44 29.3 2.0 0.425 0.031 0.017
Limestone, average 11.8 3.6 6.0 0.2 0.217 0.216
Marble, average 10.0 5.7 8.0 0.15 0.210 0.208
Silicon Carbide 8.62 2.8 17.0 0.3 0.15 0.15
Steel, AISI 304 3.45 9.9 28.0 0.29 0.12 0.11
Titanium, B120VCA 5.71 5.2 14.8 0.3 0.13 0.13
The equations of state of a linear elastic solid are
x = α (T − TO ) + σx /E u = c p (T − TO )
and
where c p , cv , α , and E are not all independent; they are related by
c p = cv + 9TO v O α2 E/(1 − 2ν)
where ν is Poisson’s ratio. Note that for iron the value of E is 28.5 × 106 psi. The
equations of state for u, h, and s should be used in difference form

not needed. This occurs because of the relative incompressibility of solids.
The mechanical equation of state is also used in difference form
= α (T2 − T1 ) + σ/E
A.6 R-12
v ft3 /lbm, u Btu/lbm, h Btu/lbm, s Btu/lbm ◦ F

In the range shown
The chord plane expression
v = vO + (T − TO ) + ( p − pO )
can be used for interpolating between values found in the table. In order to achieve
the highest accuracy, the point O should be the closest point in the table to the desired
point. In cases where there is no variation in one direction this is the equation of the
appropriate chord line, for example, in the incompressible liquid region
v P − vO
v = vO + (T − TO )
T P − TO
In the lightly shaded region, to the left and below the double line that indicates the
phase change, R-12 is an ideal gas to within 1%. In the darker shaded region it is an
ideal gas to within 0.5%. In the liquid region, above and to the right of the double
line, R-12 is incompressible to within 1%.
Source: Thermodynamic Properties of Refrigerants, R.B., Jacobsen, R.T., Penoncello,

S.G., ASHRAE, Atlanta, GA, 1986.
Appendix B
Thermodynamic Properties: SI Units

298 Appendix B: Thermodynamic Properties: SI Units
B.1 Linear Elastic Liquids
Material v0 × 103 α0 × 103 β0 × 103 cp cv

m3 /kg 1/K MPa−1 kJ/kg K kJ/kg K
Alcohol, ethyl 1.29 1.10 1.11 2.4 2.0
Benzene 1.13 1.39 0.941 1.8 1.1
Glycerin 0.784 0.504 0.222 2.4 2.2
Mercury 0.0737 0.180 0.040 0.12 0.11
Machine oil 1.09 0.738 0.767 1.7 1.5
Water 1.00 0.207 0.455 4.19 4.16
The equations of state of a linear elastic fluid are
v = v O + v O α0 (T − TO ) − v O β0 ( p − p O ) u = c p (T − TO )
and
where α0 , β0 , c p , and cv are constants listed in the table above for a few liquids,
where T0 , p0 are constant reference values. These constants are not all independent;
they satisfy the relation
c p = cv + TO v O α02 /β0
Notice that for alcohol the value of α0 is 1.10 × 10−3 K−1 , and that other values
should be read in a corresponding way. Since u, h, and s are relative quantities,
only differences have physical meaning so their equations of state should be used as
follows:

not needed. This occurs because of the relative incompressibility of liquids.
The mechanical equation of state is most often used in the incompressible form
v ∼ vO
or when specific volume differences are desired in the linear elastic form
v2 − v1 = v O α0 (T2 − T1 ) − v O β0 ( p2 − p1 )
If actual values of v are needed, the appropriate values of TO and p O are 25 ◦ C and
101 kPa.
Appendix B: Thermodynamic Properties: SI Units 299
B.2 Air as an Ideal Gas
The ideal gas law is pv = RT . Values for h, u, and s (0) are tabulated from
T T
(0)
h= c p (TC ) dTC u = h − RT s = c p (TC )/TC dTC
T0 T0
using the following expression for c p

1.5
−4 T T −1.5
c p = 1.572 · 10 + 1.339 − 15.13 +
100 100

T −2 T −3 kJ
30.59 − 21.98
100 100 kg K
for T ≥ 255.37 K and c p = 1.00276 kJ/kg K for T < 255.37 K. The reference
temperature, T0 , is 255.37 K, and the gas constant, R is 0.28701 kJ/kg K. The entropy
difference between two equilibrium states is found from the formula
s2 − s1 = s2(0) − s1(0) + R ln( p1 / p2 )
For isentropic process use the auxiliary functions
(0)
(T )−s (0) (TO )]/R T −[s (0) (T )−s (0) (TO∗ )]/R
pr (T ) = e[s or vr (T ) = e
T0∗
along with p2 / p1 = pr 2 / pr 1 or v2 /v1 = vr 2 /vr 1 At temperatures less than

100 ◦ C, air can be considered a perfect gas with a high degree of accuracy.
The equation for c p was obtained from values for N2 (75% by mass), and O2 (25% by mass) given
in: Fundamentals of Classical Thermodynamics, G.J. Van Wylen and R.E. Sonntag, John Wiley
Co., third edition SI version, New York, 1985, Table A.9, p. 652.
T Tc h u s (0) pr vr
K ◦C kJ/kg kJ/kg kJ/kg K
173.15 −100 −82.447 −132.14 −0.38963 0.25728 599.53
198.15 −75 −57.379 −114.25 −0.25439 0.41215 428.30
223.15 −50 −32.310 −96.355 −0.13525 0.62423 318.46
248.15 −25 −7.2420 −78.462 −0.028767 0.90463 244.370
273.15 0 17.894 −60.501 0.067736 1.2662 192.178
298.15 25 43.177 −42.392 0.15630 1.7239 154.07
323.15 50 68.500 −24.245 0.23786 2.2905 125.68
348.15 75 93.833 −6.0865 0.31337 2.9799 104.08
373.15 100 119.19 12.091 0.38370 3.8073 87.311
398.15 125 144.58 30.312 0.44957 4.7896 74.054
423.15 150 170.05 48.603 0.51160 5.9451 63.407
448.15 175 195.61 66.985 0.57028 7.2939 54.735
473.15 200 221.28 85.481 0.62602 8.8573 47.588
498.15 225 247.07 104.10 0.67916 10.659 41.635
523.15 250 273.01 122.87 0.72996 12.723 36.631
548.15 275 299.10 141.78 0.77867 15.076 32.290
573.15 300 325.34 160.85 0.82549 17.747 28.770
598.15 325 351.75 180.07 0.87057 20.766 25.660
623.15 350 378.31 199.46 0.91407 24.165 22.973
648.15 375 405.03 219.01 0.95611 27.977 20.638
673.15 400 431.91 238.71 0.99680 32.238 18.601
698.15 425 458.94 258.57 1.0362 36.987 16.815
723.15 450 486.14 278.59 1.0745 42.263 15.243
748.15 475 513.48 298.76 1.1117 48.107 13.854
773.15 500 540.98 319.08 1.1478 54.564 12.623
798.15 525 568.61 339.54 1.1830 61.680 11.528
823.15 550 596.40 360.15 1.2173 69.504 10.550
848.15 575 624.32 380.89 1.2507 78.085 9.6762
873.15 600 652.37 401.77 1.2833 87.477 8.8919
898.15 625 680.55 422.78 1.3151 97.738 8.1865
923.15 650 708.86 443.92 1.3462 108.92 7.5505
948.15 675 737.30 465.18 1.3766 121.08 6.9757
973.15 700 765.85 486.56 1.4063 134.30 6.4553
998.15 725 794.52 508.05 1.4354 148.62 5.9829
1023.2 750 823.30 529.66 1.4639 164.13 5.5534
1048.2 775 852.19 551.37 1.4918 180.88 5.1622
1073.2 800 881.19 573.19 1.5191 198.96 4.8050
1098.2 825 910.29 595.12 1.5459 218.44 4.4786
1123.2 850 939.49 617.14 1.5722 239.39 4.1796
1148.2 875 968.78 639.26 1.5980 261.90 3.9053
1173.2 900 998.17 661.47 1.6233 286.06 3.6534
1198.2 925 1027.6 683.78 1.6482 311.95 3.4216
T T h u s (0) pr vr
1223.2 950 1057.2 706.17 1.6726 339.66 3.2080
1248.2 975 1086.9 728.65 1.6966 369.28 3.0110
1273.2 1000 1116.6 751.21 1.7202 400.92 2.8290
1298.2 1025 1146.4 773.85 1.7434 434.66 2.6606
1323.2 1050 1176.3 796.57 1.7662 470.62 2.5046
1348.2 1075 1206.3 819.37 1.7887 508.89 2.3600
1373.2 1100 1236.3 842.24 1.8107 549.59 2.2258
1398.2 1125 1266.5 865.19 1.8325 592.84 2.1010
1423.2 1150 1296.7 888.20 1.8539 638.73 1.9849
1448.2 1175 1326.9 911.29 1.8750 687.41 1.8767
1473.2 1200 1357.2 934.44 1.8957 738.98 1.7759
1498.2 1225 1387.6 957.65 1.9162 793.57 1.6818
1523.2 1250 1418.1 980.93 1.9363 851.32 1.5939
1548.2 1275 1448.6 1004.3 1.9562 912.35 1.5116
1573.2 1300 1479.2 1027.7 1.9758 976.81 1.4347
1598.2 1325 1509.8 1051.1 1.9951 1044.8 1.3626
1623.2 1350 1540.5 1074.7 2.0142 1116.6 1.2950
1648.2 1375 1571.2 1098.2 2.0330 1192.1 1.2316
1673.2 1400 1602.1 1121.9 2.0515 1271.7 1.1720
1698.2 1425 1632.9 1145.5 2.0698 1355.5 1.1160
1723.2 1450 1663.8 1169.3 2.0879 1443.6 1.0634
1748.2 1475 1694.8 1193.1 2.1057 1536.2 1.0138
1773.2 1500 1725.8 1216.9 2.1234 1633.4 0.96707
1798.2 1525 1756.8 1240.8 2.1408 1735.4 0.92303
1823.2 1550 1788.0 1264.7 2.1579 1842.5 0.88149
1848.2 1575 1819.1 1288.7 2.1749 1954.7 0.84226
1873.2 1600 1850.3 1312.7 2.1917 2072.4 0.80521
1898.2 1625 1881.6 1336.8 2.2082 2195.5 0.77018
1923.2 1650 1912.8 1360.9 2.2246 2324.4 0.73705
1948.2 1675 1944.2 1385.1 2.2408 2459.3 0.70568
1973.2 1700 1975.6 1409.3 2.2568 2600.3 0.67598
1998.2 1725 2007.0 1433.5 2.2726 2747.7 0.64783
2023.2 1750 2038.4 1457.8 2.2883 2901.6 0.62113
2048.2 1775 2069.9 1482.1 2.3038 3062.4 0.59580
2073.2 1800 2101.5 1506.5 2.3191 3230.1 0.57176
2098.2 1825 2133.0 1530.9 2.3342 3405.1 0.54892
2123.2 1850 2164.6 1555.3 2.3492 3587.5 0.52722
2148.2 1875 2196.3 1579.8 2.3640 3777.6 0.50658
2173.2 1900 2228.0 1604.3 2.3787 3975.6 0.48695
2198.2 1925 2259.7 1628.8 2.3932 4181.9 0.42826
2223.2 1950 2291.5 1653.4 2.4075 4396.5 0.45047
2248.2 1975 2323.5 1678.0 2.4218 4619.8 0.43351
T T h u s (0) pr vr
2273.2 2000 2355.1 1702.7 2.4358 4852.1 0.41735
2298.2 2025 2386.9 1727.4 2.4498 5093.6 0.40193
2323.2 2050 2418.8 1752.1 2.4636 5344.5 0.38723
2348.2 2075 2450.8 1776.8 2.4772 5605.2 0.37319
2373.2 2100 2482.7 1801.6 2.4908 5876.0 0.35979
2398.2 2125 2514.7 1826.4 2.5042 6157.0 0.34698
2423.2 2150 2546.7 1851.3 2.5175 6448.7 0.33474
2448.2 2175 2578.8 1876.1 2.5306 6751.2 0.32304
2473.2 2200 2610.9 1901.1 2.5437 7065.0 0.31185
2498.2 2225 2643.0 1926.0 2.5566 7390.3 0.30113
2523.2 2250 2675.1 1951.0 2.5694 7727.4 0.29088
2548.2 2275 2707.3 1975.9 2.5821 8076.6 0.28106
2573.2 2300 2739.5 2001.1 2.5947 8438.3 0.27165
2598.2 2325 2771.7 2026.0 2.6071 8812.8 0.26263
2623.2 2350 2804.0 2051.1 2.6195 9200.5 0.25399
2648.2 2375 2836.2 2076.2 2.6317 9601.6 0.24570
2673.2 2400 2868.5 2101.3 2.6439 10017 0.23774
2698.2 2425 2900.9 2126.5 2.6559 10446 0.23011
2723.2 2450 2933.2 2151.7 2.6678 10889 0.23011
2748.2 2475 2965.6 2176.9 2.6797 11348 0.21574
2773.2 2500 2998.0 2202.1 2.6914 11822 0.20897
2798.2 2525 3030.5 2227.4 2.7031 12311 0.20247
2823.2 2550 3062.9 2252.7 2.7146 12817 0.19622
2848.2 2575 3095.4 2278.0 2.7261 13339 0.19021
2873.2 2600 3127.9 2303.3 2.7374 13878 0.18443
2898.2 2625 3160.5 2328.7 2.7487 14434 0.17887
2923.2 2650 3193.0 2354.1 2.7599 15008 0.17351
2948.2 2675 3225.6 2379.5 2.7710 15600 0.16836
2973.2 2700 3258.2 2404.9 2.7820 16210 0.16339
2998.2 2725 3290.9 2430.4 2.7930 16839 0.15861
3023.2 2750 3323.5 2455.9 2.8038 17488 0.15400
3048.2 2775 3356.2 2481.4 2.8146 18156 0.14956
3073.2 2800 3388.9 2506.9 2.8253 18845 0.14527
3098.2 2825 3421.6 2532.5 2.8359 19554 0.14114
3123.2 2850 3454.4 2558.0 2.8464 20285 0.13716
3148.2 2875 3487.2 2583.6 2.8568 21037 0.13331
3173.2 2900 3520.0 2609.3 2.8672 21812 0.12960
3198.2 2925 3552.8 2634.9 2.8775 22609 0.12602
3223.2 2950 3585.6 2660.6 2.8877 23429 0.12256
3248.2 2975 3618.5 2686.2 2.8979 24273 0.11921
3273.2 3000 3651.3 2711.9 2.9080 25140 0.11598
B.3 Perfect Gases
The equations of state of a perfect gas are
R
pv = RT u = cv T = T
k−1
and
k
h = cpT = RT s = R ln[T 1/(k−1) v]
k−1
where R, c p , and cv are constants listed in the table, and k is the ratio of specific heats
c p /cv . The constants are not all independent, they satisfy the relation c p = cv + R.
u 2 − u 1 = cv (T2 − T1 ) h 2 − h 1 = c p (T2 − T1 ) s2 − s1 = R ln[(T2 /T1 )1/(k−1) (v2 /v1 )]
Two other expressions of the entropic equation of state, which are sometimes useful,
can be obtained by eliminating either v or T using the ideal gas law
s = R ln R + R ln[T k/(k−1) / p] s = −cv ln R + cv ln[ pv k ]
They are also used in difference form
s2 − s1 = R ln[(T2 /T1 )k/(k−1) ( p1 / p2 )] s2 − s1 = cv ln[( p2 / p1 )(v2 /v1 )k ]
In addition, it is often desirable to use the mechanical equation of state in dimen-

sionless form p2 v2 /T2 = p1 v1 /T1
A gas can be considered perfect only over a limited temperature range. Here the
(constant) values of the specific heats correspond to the vicinity of 25 ◦ C.
B.4 Water
v m3 /kg, u kJ/kg, h kJ/kg, s kJ/kg ◦ C

In the range shown
In the shaded region correspond- In the shaded region corresponding

ing to an incompressible liquid (this to an ideal gas (this appears in the
appears in the table to the right and table to the left and below the double
above the double line indicating the line indicating the phase change), the
phase change), the following rela- following relations hold, to within
tions hold, to within 1%, on a row 1%, on a row (at constant T ):
(at constant T ):
v2 = p1 v1 / p2
v = vf u2 = u1
u = uf h2 = h1
h = h f + ( p − ps )v f s2 = s1 + R ln( p1 / p2 )
s = sf
The value of R is 0.4614 kJ/kg ◦ C,
The subscripted values are those of the lightest shading is 1%, and the
the saturated liquid; they also appear darkest is 0.1%.
to the right of the phase change line.
Source: Steam Tables, J.H. Keenan, F.G. Keyes, P.G. Hill, J.G. Moore, John Wiley and Sons, New
York, 1969.
B.5 Linear Elastic Solids
Material v0 × 103 αl × 106 E ν cp cv

m3 /kg 1/K GPa kJ/kg K kJ/kg K
Aluminum, 2024-T3 0.361 22.7 73.1 0.33 0.96 0.88
Copper 0.112 16.6 110 0.345 0.38 0.36
Glass, average 0.320 7.7 63.4 0.235 0.833 0.829
Iron 0.127 11.7 196 0.27 0.452 0.435
Lead, pure 0.0882 52.7 13.8 0.425 0.13 0.071
Limestone, average 0.426 6.5 41.4 0.2 0.909 0.904
Marble, average 0.361 10.3 55.2 0.15 0.879 0.871
Silicon Carbide 0.311 5.0 117 0.3 0.63 0.63
Steel, AISI 304 0.125 17.8 193 0.29 0.50 0.46
Titanium, B120VCA 0.206 9.4 102 0.3 0.54 0.54
The equations of state of a linear elastic solid are
x = α (T − TO ) + σx /E u = c p (T − TO )
and
where c p , cv , α , and E are not all independent; they are related by
c p = cv + 9TO v O α2 E/(1 − 2ν)
where ν is Poisson’s ratio. Note that for iron the value of v0 is 0.127 × 10−3 m3 /kg.

not needed. This occurs because of the relative incompressibility of solids.
The mechanical equation of state is also used in difference form
= α (T2 − T1 ) + σ/E
B.6 R-12
v m3 /kg, u kJ/kg, h kJ/kg, s kJ/kg ◦ C

In the range shown
The chord plane expression
v = vO + (T − TO ) + ( p − pO )
can be used for interpolating between values found in the table. In order to achieve
the highest accuracy, the point O should be the closest point in the table to the desired
point. In cases where there is no variation in one direction this is the equation of the
appropriate chord line, for example, in the incompressible liquid region
v P − vO
v = vO + (T − TO )
T P − TO
In the lightly shaded region, to the left and below the double line that indicates the
phase change, R-12 is an ideal gas to within 1%. In the darker shaded region, it is an
ideal gas to within 0.5%. In the liquid region, above and to the right of the double
line, R-12 is incompressible to within 1%.
Source: Thermodynamic Properties of Refrigerants, Stewart, R.B., Jacobsen, R.T., Penoncello, S.G.,
ASHRAE, Atlanta, GA, 1986.
Index
A losses, 263
Absolute quantity, 9 regeneration, 261
Acceleration, 16 reheat, 259
Additive, 2, 12 British Thermal Unit, 118, 134
Adiabatic, 32, 112 Bulk modulus, 42
process, 123
Air conditioner, 216
Air Standard Cycle C
Otto, 252 Calorimetric process, 119
Air standard power cycles, 251 Calorimetry, 119
Avogadro Calory, 134
Amadeo, 72 Carnot
constant, 72 cycle, 212, 213, 220
law, 72 efficiency, 213, 219
engine, 213, 219
heat pump, 216
B refrigerator, 218, 267
Back work ratio, 257 Sadi, 212
Barometer, 28 theorem, 209
Beau de Rochas Clapeyron’s equation, 215
Alphonse, 251 Clausius
Boiler, 233, 236, 238, 241 Rudolf, 130, 176, 194
pipe strength, 245 theorem, 205
pressure, 242, 244 Clearance volume, 251
Boiling, 68 Coefficient of performance, 211, 216, 218,
temperature, 68 267
Boltzmann Coefficient of thermal expansion, 41
constant, 40, 71 ideal gas, 74
Ludwig, 40, 178, 179 Combustion, 252
Bottom dead center, 251 Compressibility, 42
Bound ideal gas, 74
heating, 204 isothermal, 42
power, 207 Compression ratio, 252
Brayton Compressor, 157, 261, 264, 268
George, B., 257 loss, 263, 270
Brayton cycle, 256 pressure ratio, 257
intercooling, 259 Condenser, 237, 241, 268
316 Index
pressure, 239, 242, 244 time, 2

Condensible gas velocity, 15
energy, 149 volume, 11
enthalpy, 148 weight, 12
entropy, 189 work, 24
volume, 76 Displacement, 233, 251
Conduction, 114 Duct flow, 165
Conservation
of energy, 132, 154
of heating, 113 E
of mass, 12, 137, 153, 155 Effectiveness, 221
of power, 94 Efficiency
Continuous Carnot, 213, 219
process, 132 compressor, 223, 263
Convection, 114 mechanical, 99, 103, 233
Crankshaft, 251 pump, 250
Critical point, 76 regenerator, 261
Cut off thermal, 211, 241
steam, 239 thermodynamic, 219, 220, 222, 267
Cutoff ratio, 254 Rankine cycle, 243
Cycle, 106 Watt engine, 243
aircraft gas turbine, 264 turbine, 223, 250, 263
air standard, 251 Energetic equation of state, 138
Brayton, 256 condensible gas, 149
Diesel, 254 ideal gas, 144
Otto, 252 liquids and solids, 140
Rankine, see Rankine Cycle liquid-vapor equilibrium, 150
vapor refrigeration, 267 perfect gas, 147
Cyclic Energy
process, 132 balance, 134, 155
Cylinder, 251 internal, 130, 131
kinetic, 23
transfer, 131
D Energy equation
Density, 15 closed system, 131
power, 252 filling tank, 168
Diesel open system, 154
Rudolf, 254 steady flow, 155
Diffuser, 165, 264 Engine
Dimension, 2 bore, 253
acceleration, 17 displacement, 233
area, 6 double acting, 241
energy, 24 four stroke, 252
entropy, 176 heat, 211, 231, 266
force, 17 jet, 264
heat transfer, 131 thrust, 264
length, 2 knock, 254
mass, 12 mep, 252
momentum, 23 Newcomen, 231–235
power, 89 piston cylinder, 256
pressure, 25 power, 253
quantity, 71 speed, 241
specific volume, 14 steam, 237
Index 317
stroke, 253 Willard, J., 84, 179

torque, 253
Watt, 237–241
Enthalpic equation of state, 139 H
condensible gas, 148 Heat absorption, 116
ideal gas, 144 Heat engine, 211
liquids and solids, 140 Heat exchanger, 163
liquid-vapor equilibrium, 150 counterflow, 261
perfect gas, 147 Heat flux vector, 112
Enthalpy, 137 Heat pump, 266
Entropic equation of state, 179, 181 Heat transfer, 112
condensible gas, 189 constraint, 204
ideal gases, 185 modes, 114
liquids and solids, 182 Heat transfer rate, 112
liquid–vapor equilibrium, 191 Heating, 112
perfect gas, 187 bound, 204
vaporizable liquids, 183 Horsepower, 90
Entropy, 175, 176 Hotness, 32–35, 38, 115
balance, 202 ideal gas, see Temperature, ideal gas
generation, 194, 195, 198, 202
kinetic, 178
transfer, 193 I
Equilibrium Ideal gas
liquid-vapor, 81 energy, 144
macroscopic, 50, 67, 78 enthalpy, 144
material, 84, 192 entropy, 185
phase, 80, 84 volume, 72
process, 105, 106, 138, 179 Impulse, 23
state, 50, 51 Intensive, 15
static, 47 Intercooling, 259
thermal, 47 Interpolation, 54
thermodynamic, 47, 50 Irreversibility principle, 192
work, 106 Irreversible
Evaporator, 268 process, 194, 202
Extensive, 12 Isentropic
compression, 242
expansion, 242
F process, 177, 183, 239, 252
Feedwater heater, 247 Isobaric
First law of thermodynamics, 130 curve, 51
Fourier process, 108, 109, 137
Joseph, 114 Isochoric
law of conduction, 114 curve, 51
process, 123, 252
Isothermal
G curve, 51
Gas constant, 72 process, 110, 123, 177
Geometry
of curves, 52–55
of surfaces, 55–60 J
Gibbs Joule
free energy, 192 experiment, 133, 143
phase rule, 84 James P., 133, 143
318 Index
K law of universal gravitation, 17

Kelvin, see Thomson, William Nozzle, 165, 264
-Plank corollary, 208
Kilocalorie, 118, 134
O
Otto
L Nikolas, 252
Latent heat
constant pressure, 116
P
constant volume, 116
Pascal, 25
Liquid
Perfect gas
energy, 140
energy, 147
enthalpy, 140
enthalpy, 147
entropy, 182
entropy, 187
volume, 66
Perpetual motion, 132, 208
Liquid–vapor equilibrium
Phase rule, 84
energy, 150
Piston, 251
enthalpy, 150
Poisson’s ratio, 67
entropy, 191
Polytropic
volume, 81
change of state, 74
process, 110
Power, 231
M
average, 90
Manometer, 27, 28
bound, 207
Mass, 12
density, 252
balance, 155
Power plant
Maxwell
gas turbine, 257
James Clerk, 179, 180
steam, 237
relation, 180, 182, 188
Preheat, 261
Mayer
Pre-ignition, 254
J.Robert, 130
Pressure, 24
Mechanical equation of state, 50, 64
atmospheric, 28
condensible gases, 76
extraction, 247
ideal gases, 72
gauge, 28
liquids, 66
kinetic, 31
liquid-vapor equilibrium, 81
mean effective, 238, 252, 253
solids, 67
saturation, 68
vaporizable liquids, 68
Prime mover, 233, 236, 237, 242
Mixing chamber, 162
Property, see Dimension
Molar mass, 72
Pump, 157, 231, 237, 242, 247
Molar volume, 72
Mole, 72
Momentum, 23 Q
Quality, 81
Quasi-equilibrium
N process, 106, 108
Newcomen state, 50
engine, see Engine
Thomas, 231
Newton R
Isaac, 17 Rankine
law of action and reaction, 94 William Macquorn, 130, 176, 241
law of cooling, 114 Rankine cycle, 241–250
law of motion, 17 ideal, 242
Index 319
regeneration, 247 English, 281

reheat, 244, 245 SI, 299
superheat, 244 liquids
turbine loss, 249 English, 280
Ratio SI, 298
back work, 257 perfect gases
Refrigerating capacity, 267 English, 285
Refrigeration, 266 SI, 303
Ton, 266 R-12
Refrigerator, 266 English, 292
Carnot, 267 SI, 310
Regemerator, 261 solids
Regeneration, 259, 261 English, 291
Rankine cycle, 247 SI, 309
Reheat, 259 water
Relative quantity, 9 English, 286
Reservoir, 120 SI, 304
Reversible Temperature, 32
process, 194, 202, 213, 219 celsius, 33
fahrenheit, 33
gas, 37
S ideal gas, 38, 185
Saturated liquid curve, 68 kelvin scale, 38
Saturated vapor curve, 76 kinetic, 40
Saturation liquid, 35
curve, 68, 76 rankine scale, 38
pressure, 68 saturation, 68
Second law of thermodynamics, 194, 201, thermodynamic, 177, 214
269 Thermal mass, 120, 200
Simple compressible substances, 50 Thomson
Solid William, 130, 194
energy, 140 Throttle, 241, 247, 267
enthalpy, 140 Throttling process, 166
entropy, 182 Top dead center, 251
strain, 67
Traction, 24
Specific fuel consumption, 236, 241
Transitivity, 2
Specific heat
Trap, 247
constant pressure, 116, 119
Triple point, 38, 84
constant volume, 116, 121
Turbine, 157, 264
Specific volume, 14
blade erosion, 244
Steady flow, 256
gas, 256, 261
Steady state
loss, 249, 263
process, 132
pressure ratio, 257
Steam
steam, 242
expansive power, 239, 245
Subcooling
vapor refrigeration, 269
Superheating U
vapor refrigeration, 269 Universal gas constant, 71
T V
Table Valve
air exhaust, 251
320 Index
intake, 251 engine, 237–241

Vaporizable liquid James, 90, 237, 239, 251
energy, 142 Weight, 12
enthalpy, 141 Work, 23–24, 89
entropy, 183 compressor, 259
volume, 68 constraint, 207
Vapor refrigeration electrical, 160
compressor loss, 270 equilibrium, 106, 108
cycle, 267–271 flow, 154
subcooling, 269 reversible, 91, 208
superheating, 269 stirring, 102
Velocity, 15
of sound, 106
Volume, 12 Y
clearance, 251 Young’s modulus, 67
specific, 14, 15
W Z
Watt, 90, 160 Zeroth law of thermodynamics, 35

Thermodynamics Basic Principles

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Thermodynamics Basic Principles

Hochgeladen von

Copyright:

Verfügbare Formate

Mechanical Engineering Series

More information about this series at http://www.springer.com/series/1161

Additional material to this book can be downloaded from http://extras.springer.com.

ISSN 0941-5122 ISSN 2192-063X (electronic)

In this text, I have presented thermodynamics as mechanics, fluid mechanics, and

also used in discussing equations of state. In Chap. 2, the mechanical equation of

Villanova, PA, USA Alan M. Whitman

1 Measurement and Properties of Matter . . . . . . . . . . . . . . . . . . . . . . 1

2.4 The Mechanical Equation of State . . . . . . . . . . . . . . . . . . . . . . . . 64

5.3.1 Liquids and Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

Appendix A: Thermodynamic Properties: English Units . . . . . . . . . . . . . 279

© Springer Nature Switzerland AG 2020 1

experimentally determined equations of state. Usually this viewpoint, which is the

1.2 Dimensions and Units

L0 < L1 < L2 < L3 < L4

where we have used t = ts = 1 d, and tj = 1 s. We make these replacements because

Example 1.1 If L1 , and L2 are lengths, and t1 , and t2 are times,

is not a valid equation, because it is not dimensionally homogeneous. The

is valid because it is dimensionally homogeneous. The composite dimension

is valid because it is dimensionally homogeneous. The composite dimension

which we use as we indicated previously

When it is in unitwise homogeneous form the terms add numerically and

One way to ensure unitwise homogeneity is to express all quantities in an equation

1.2.1 Fundamental and Derived Dimensions

On cross-multiplying, this takes the form

or written as a dimensional equation involving only one square

kS = 2.296 · 10−5 acres/ft2 (1.8)

A = (2.296 · 10−5 acres/ft 2 )L2

2. The constant kS is taken as a pure (dimensionless) number. Accordingly the dimen-

2.296 · 10−5 acres = 1 ft2 (1.10)

0 0.5 1.5 2 in Distance

2.54 3.81 cm from O

1.2.2 Absolute and Relative Quantities

ζA − ζB = xA/B = xA/O − xB/O = xA/O − xB/O

xo = xo + xO /O (1.11)

Table 1.1 Approximate mileage between major eastern seaboard cities

Big Creation of Birth of Flight of

ta = toj + tj = toj + 3761yr + tc = toj + 4383yr + ti

1.3 Properties of Matter

Matter we recognize as existing in various gaseous, liquid, or solid forms. In any of

1 gal = 0.1337 ft3 (1.12)

We call volume an extensive, meaning additive, property. Since we consider matter

1.3.2 Weight and Mass

where ke is a dimensional constant that characterizes the earth, with dimensions

1.3.3 Density and Specific Volume

which for continuous distributions of matter is an integral

1.3.4 Velocity and Acceleration

xP/O (t2 ) − xP/O (t1 )

which we call the velocity of the mass center.

1 kgf = 9.807 kg m/s2

1 slug = 32.17 lbm

1 lbf = 1 slug ft/s2 (1.31)

however we do not measure xg explicitly, rather we introduce a linear force measure

allow us to solve for the constants C1 and C2 to obtain

From this we find that

and on substituting into Eq. (1.32), Fx = WG (xg − x0 )/(xG − x0 ),

In the accompanying figure we plot, in dimensionless form, the difference

εWG2 = (0.01 lbf −2 )(100 lbf 2 ) = 1

Then, if 5 lbf is exerted

The largest error in the range

1.3.6 Impulse and Momentum

1.3.7 Work and Energy

Another integral of the equation of motion of a particle, or mass center of a rigid

ζA − ζB = xA/B = xA/O − xB/O = xA/O − xB/O

xo = xo + xO /O (1.11)